title
stringlengths 15
153
| url
stringlengths 97
97
| authors
stringlengths 6
328
| detail_url
stringlengths 97
97
| tags
stringclasses 1
value | Bibtex
stringlengths 54
54
⌀ | Paper
stringlengths 93
93
⌀ | Reviews And Public Comment »
stringlengths 63
65
⌀ | Supplemental
stringlengths 100
100
⌀ | abstract
stringlengths 310
2.42k
⌀ | Supplemental Errata
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|
Towards a Theoretical Framework of Out-of-Distribution Generalization
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5c1cb0bebd56ae38817b251ad72bedb-Abstract.html
|
Haotian Ye, Chuanlong Xie, Tianle Cai, Ruichen Li, Zhenguo Li, Liwei Wang
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5c1cb0bebd56ae38817b251ad72bedb-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13424-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c5c1cb0bebd56ae38817b251ad72bedb-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=kFJoj7zuDVi
|
https://papers.nips.cc/paper_files/paper/2021/file/c5c1cb0bebd56ae38817b251ad72bedb-Supplemental.pdf
|
Generalization to out-of-distribution (OOD) data is one of the central problems in modern machine learning. Recently, there is a surge of attempts to propose algorithms that mainly build upon the idea of extracting invariant features. Although intuitively reasonable, theoretical understanding of what kind of invariance can guarantee OOD generalization is still limited, and generalization to arbitrary out-of-distribution is clearly impossible. In this work, we take the first step towards rigorous and quantitative definitions of 1) what is OOD; and 2) what does it mean by saying an OOD problem is learnable. We also introduce a new concept of expansion function, which characterizes to what extent the variance is amplified in the test domains over the training domains, and therefore give a quantitative meaning of invariant features. Based on these, we prove an OOD generalization error bound. It turns out that OOD generalization largely depends on the expansion function. As recently pointed out by Gulrajani & Lopez-Paz (2020), any OOD learning algorithm without a model selection module is incomplete. Our theory naturally induces a model selection criterion. Extensive experiments on benchmark OOD datasets demonstrate that our model selection criterion has a significant advantage over baselines.
| null |
Slice Sampling Reparameterization Gradients
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5c3d4fe6b2cc463c7d7ecba17cc9de7-Abstract.html
|
David Zoltowski, Diana Cai, Ryan P. Adams
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5c3d4fe6b2cc463c7d7ecba17cc9de7-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13425-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c5c3d4fe6b2cc463c7d7ecba17cc9de7-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=X4_aAfxsOoE
|
https://papers.nips.cc/paper_files/paper/2021/file/c5c3d4fe6b2cc463c7d7ecba17cc9de7-Supplemental.pdf
|
Many probabilistic modeling problems in machine learning use gradient-based optimization in which the objective takes the form of an expectation. These problems can be challenging when the parameters to be optimized determine the probability distribution under which the expectation is being taken, as the na\"ive Monte Carlo procedure is not differentiable. Reparameterization gradients make it possible to efficiently perform optimization of these Monte Carlo objectives by transforming the expectation to be differentiable, but the approach is typically limited to distributions with simple forms and tractable normalization constants. Here we describe how to differentiate samples from slice sampling to compute \textit{slice sampling reparameterization gradients}, enabling a richer class of Monte Carlo objective functions to be optimized. Slice sampling is a Markov chain Monte Carlo algorithm for simulating samples from probability distributions; it only requires a density function that can be evaluated point-wise up to a normalization constant, making it applicable to a variety of inference problems and unnormalized models. Our approach is based on the observation that when the slice endpoints are known, the sampling path is a deterministic and differentiable function of the pseudo-random variables, since the algorithm is rejection-free. We evaluate the method on synthetic examples and apply it to a variety of applications with reparameterization of unnormalized probability distributions.
| null |
Multi-Label Learning with Pairwise Relevance Ordering
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5d215777c229704a7862de577d40a73-Abstract.html
|
Ming-Kun Xie, Sheng-Jun Huang
|
https://papers.nips.cc/paper_files/paper/2021/hash/c5d215777c229704a7862de577d40a73-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13426-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c5d215777c229704a7862de577d40a73-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=ICBPhB079dQ
|
https://papers.nips.cc/paper_files/paper/2021/file/c5d215777c229704a7862de577d40a73-Supplemental.pdf
|
Precisely annotating objects with multiple labels is costly and has become a critical bottleneck in real-world multi-label classification tasks. Instead, deciding the relative order of label pairs is obviously less laborious than collecting exact labels. However, the supervised information of pairwise relevance ordering is less informative than exact labels. It is thus an important challenge to effectively learn with such weak supervision. In this paper, we formalize this problem as a novel learning framework, called multi-label learning with pairwise relevance ordering (PRO). We show that the unbiased estimator of classification risk can be derived with a cost-sensitive loss only from PRO examples. Theoretically, we provide the estimation error bound for the proposed estimator and further prove that it is consistent with respective to the commonly used ranking loss. Empirical studies on multiple datasets and metrics validate the effectiveness of the proposed method.
| null |
Sampling with Trusthworthy Constraints: A Variational Gradient Framework
|
https://papers.nips.cc/paper_files/paper/2021/hash/c61aed648da48aa3893fb3eaadd88a7f-Abstract.html
|
Xingchao Liu, Xin Tong, Qiang Liu
|
https://papers.nips.cc/paper_files/paper/2021/hash/c61aed648da48aa3893fb3eaadd88a7f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13427-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c61aed648da48aa3893fb3eaadd88a7f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=B2cyX_ht4VI
|
https://papers.nips.cc/paper_files/paper/2021/file/c61aed648da48aa3893fb3eaadd88a7f-Supplemental.pdf
|
Sampling-based inference and learning techniques, especially Bayesian inference, provide an essential approach to handling uncertainty in machine learning (ML). As these techniques are increasingly used in daily life, it becomes essential to safeguard the ML systems with various trustworthy-related constraints, such as fairness, safety, interpretability. Mathematically, enforcing these constraints in probabilistic inference can be cast into sampling from intractable distributions subject to general nonlinear constraints, for which practical efficient algorithms are still largely missing. In this work, we propose a family of constrained sampling algorithms which generalize Langevin Dynamics (LD) and Stein Variational Gradient Descent (SVGD) to incorporate a moment constraint specified by a general nonlinear function. By exploiting the gradient flow structure of LD and SVGD, we derive two types of algorithms for handling constraints, including a primal-dual gradient approach and the constraint controlled gradient descent approach. We investigate the continuous-time mean-field limit of these algorithms and show that they have O(1/t) convergence under mild conditions. Moreover, the LD variant converges linearly assuming that a log Sobolev like inequality holds. Various numerical experiments are conducted to demonstrate the efficiency of our algorithms in trustworthy settings.
| null |
Robust and Decomposable Average Precision for Image Retrieval
|
https://papers.nips.cc/paper_files/paper/2021/hash/c622c085c04eadc473f08541b255320e-Abstract.html
|
Elias Ramzi, Nicolas THOME, Clément Rambour, Nicolas Audebert, Xavier Bitot
|
https://papers.nips.cc/paper_files/paper/2021/hash/c622c085c04eadc473f08541b255320e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13428-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c622c085c04eadc473f08541b255320e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=VjQw3v3FpJx
|
https://papers.nips.cc/paper_files/paper/2021/file/c622c085c04eadc473f08541b255320e-Supplemental.pdf
|
In image retrieval, standard evaluation metrics rely on score ranking, e.g. average precision (AP). In this paper, we introduce a method for robust and decomposable average precision (ROADMAP) addressing two major challenges for end-to-end training of deep neural networks with AP: non-differentiability and non-decomposability.Firstly, we propose a new differentiable approximation of the rank function, which provides an upper bound of the AP loss and ensures robust training. Secondly, we design a simple yet effective loss function to reduce the decomposability gap between the AP in the whole training set and its averaged batch approximation, for which we provide theoretical guarantees.Extensive experiments conducted on three image retrieval datasets show that ROADMAP outperforms several recent AP approximation methods and highlight the importance of our two contributions. Finally, using ROADMAP for training deep models yields very good performances, outperforming state-of-the-art results on the three datasets.Code and instructions to reproduce our results will be made publicly available at https://github.com/elias-ramzi/ROADMAP.
| null |
Fast rates for prediction with limited expert advice
|
https://papers.nips.cc/paper_files/paper/2021/hash/c688defd45ad6638febd469adb09ddf7-Abstract.html
|
El Mehdi Saad, Gilles Blanchard
|
https://papers.nips.cc/paper_files/paper/2021/hash/c688defd45ad6638febd469adb09ddf7-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13429-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c688defd45ad6638febd469adb09ddf7-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=7m6qvNqFjr
|
https://papers.nips.cc/paper_files/paper/2021/file/c688defd45ad6638febd469adb09ddf7-Supplemental.pdf
|
We investigate the problem of minimizing the excess generalization error with respect to the best expert prediction in a finite family in the stochastic setting, under limited access to information. We consider that the learner has only access to a limited number of expert advices per training round, as well as for prediction. Assuming that the loss function is Lipschitz and strongly convex, we show that if we are allowed to see the advice of only one expert per round in the training phase, or to use the advice of only one expert for prediction in the test phase, the worst-case excess risk is ${\Omega}(1/\sqrt{T})$ with probability lower bounded by a constant. However, if we are allowed to see at least two actively chosen expert advices per training round and use at least two experts for prediction, the fast rate $\mathcal{O}(1/T)$ can be achieved. We design novel algorithms achieving this rate in this setting, and in the setting where the learner have a budget constraint on the total number of observed experts advices, and give precise instance-dependent bounds on the number of training rounds needed to achieve a given generalization error precision.
| null |
Probabilistic Transformer For Time Series Analysis
|
https://papers.nips.cc/paper_files/paper/2021/hash/c68bd9055776bf38d8fc43c0ed283678-Abstract.html
|
Binh Tang, David S Matteson
|
https://papers.nips.cc/paper_files/paper/2021/hash/c68bd9055776bf38d8fc43c0ed283678-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13430-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c68bd9055776bf38d8fc43c0ed283678-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=HfpNVDg3ExA
|
https://papers.nips.cc/paper_files/paper/2021/file/c68bd9055776bf38d8fc43c0ed283678-Supplemental.pdf
|
Generative modeling of multivariate time series has remained challenging partly due to the complex, non-deterministic dynamics across long-distance timesteps. In this paper, we propose deep probabilistic methods that combine state-space models (SSMs) with transformer architectures. In contrast to previously proposed SSMs, our approaches use attention mechanism to model non-Markovian dynamics in the latent space and avoid recurrent neural networks entirely. We also extend our models to include several layers of stochastic variables organized in a hierarchy for further expressiveness. Compared to transformer models, ours are probabilistic, non-autoregressive, and capable of generating diverse long-term forecasts with uncertainty estimates. Extensive experiments show that our models consistently outperform competitive baselines on various tasks and datasets, including time series forecasting and human motion prediction.
| null |
A Hierarchical Reinforcement Learning Based Optimization Framework for Large-scale Dynamic Pickup and Delivery Problems
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6a01432c8138d46ba39957a8250e027-Abstract.html
|
Yi Ma, Xiaotian Hao, Jianye Hao, Jiawen Lu, Xing Liu, Tong Xialiang, Mingxuan Yuan, Zhigang Li, Jie Tang, Zhaopeng Meng
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6a01432c8138d46ba39957a8250e027-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13431-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c6a01432c8138d46ba39957a8250e027-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=2F_wnaioS6
|
https://papers.nips.cc/paper_files/paper/2021/file/c6a01432c8138d46ba39957a8250e027-Supplemental.pdf
|
The Dynamic Pickup and Delivery Problem (DPDP) is an essential problem in the logistics domain, which is NP-hard. The objective is to dynamically schedule vehicles among multiple sites to serve the online generated orders such that the overall transportation cost could be minimized. The critical challenge of DPDP is the orders are not known a priori, i.e., the orders are dynamically generated in real-time. To address this problem, existing methods partition the overall DPDP into fixed-size sub-problems by caching online generated orders and solve each sub-problem, or on this basis to utilize the predicted future orders to optimize each sub-problem further. However, the solution quality and efficiency of these methods are unsatisfactory, especially when the problem scale is very large. In this paper, we propose a novel hierarchical optimization framework to better solve large-scale DPDPs. Specifically, we design an upper-level agent to dynamically partition the DPDP into a series of sub-problems with different scales to optimize vehicles routes towards globally better solutions. Besides, a lower-level agent is designed to efficiently solve each sub-problem by incorporating the strengths of classical operational research-based methods with reinforcement learning-based policies. To verify the effectiveness of the proposed framework, real historical data is collected from the order dispatching system of Huawei Supply Chain Business Unit and used to build a functional simulator. Extensive offline simulation and online testing conducted on the industrial order dispatching system justify the superior performance of our framework over existing baselines.
| null |
Spatio-Temporal Variational Gaussian Processes
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6b8c8d762da15fa8dbbdfb6baf9e260-Abstract.html
|
Oliver Hamelijnck, William Wilkinson, Niki Loppi, Arno Solin, Theodoros Damoulas
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6b8c8d762da15fa8dbbdfb6baf9e260-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13432-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c6b8c8d762da15fa8dbbdfb6baf9e260-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=rJhCP_vC6T
|
https://papers.nips.cc/paper_files/paper/2021/file/c6b8c8d762da15fa8dbbdfb6baf9e260-Supplemental.pdf
|
We introduce a scalable approach to Gaussian process inference that combines spatio-temporal filtering with natural gradient variational inference, resulting in a non-conjugate GP method for multivariate data that scales linearly with respect to time. Our natural gradient approach enables application of parallel filtering and smoothing, further reducing the temporal span complexity to be logarithmic in the number of time steps. We derive a sparse approximation that constructs a state-space model over a reduced set of spatial inducing points, and show that for separable Markov kernels the full and sparse cases exactly recover the standard variational GP, whilst exhibiting favourable computational properties. To further improve the spatial scaling we propose a mean-field assumption of independence between spatial locations which, when coupled with sparsity and parallelisation, leads to an efficient and accurate method for large spatio-temporal problems.
| null |
MERLOT: Multimodal Neural Script Knowledge Models
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6d4eb15f1e84a36eff58eca3627c82e-Abstract.html
|
Rowan Zellers, Ximing Lu, Jack Hessel, Youngjae Yu, Jae Sung Park, Jize Cao, Ali Farhadi, Yejin Choi
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6d4eb15f1e84a36eff58eca3627c82e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13433-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c6d4eb15f1e84a36eff58eca3627c82e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=CRFSrgYtV7m
|
https://papers.nips.cc/paper_files/paper/2021/file/c6d4eb15f1e84a36eff58eca3627c82e-Supplemental.pdf
|
As humans, we understand events in the visual world contextually, performing multimodal reasoning across time to make inferences about the past, present, and future. We introduce MERLOT, a model that learns multimodal script knowledge by watching millions of YouTube videos with transcribed speech -- in an entirely label-free, self-supervised manner. By pretraining with a mix of both frame-level (spatial) and video-level (temporal) objectives, our model not only learns to match images to temporally corresponding words, but also to contextualize what is happening globally over time. As a result, MERLOT exhibits strong out-of-the-box representations of temporal commonsense, and achieves state-of-the-art performance on 12 different video QA datasets when finetuned. It also transfers well to the world of static images, allowing models to reason about the dynamic context behind visual scenes. On Visual Commonsense Reasoning, MERLOT~answers questions correctly with 80.6\% accuracy, outperforming state-of-the-art models of similar size by over 3\%, even those that make heavy use of auxiliary supervised data (like object bounding boxes).Ablation analyses demonstrate the complementary importance of: 1) training on videos versus static images; 2) scaling the magnitude and diversity of the pretraining video corpus; and 3) using diverse objectives that encourage full-stack multimodal reasoning, from the recognition to cognition level.
| null |
Fast Approximate Dynamic Programming for Infinite-Horizon Markov Decision Processes
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6f798b844366ccd65d99bc7f31e0e02-Abstract.html
|
Mohamad Amin Sharifi Kolarijani, Gyula Max, Peyman Mohajerin Mohajerin Esfahani
|
https://papers.nips.cc/paper_files/paper/2021/hash/c6f798b844366ccd65d99bc7f31e0e02-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13434-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c6f798b844366ccd65d99bc7f31e0e02-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=ZW-ZsleMIWk
|
https://papers.nips.cc/paper_files/paper/2021/file/c6f798b844366ccd65d99bc7f31e0e02-Supplemental.pdf
|
In this study, we consider the infinite-horizon, discounted cost, optimal control of stochastic nonlinear systems with separable cost and constraints in the state and input variables. Using the linear-time Legendre transform, we propose a novel numerical scheme for implementation of the corresponding value iteration (VI) algorithm in the conjugate domain. Detailed analyses of the convergence, time complexity, and error of the proposed algorithm are provided. In particular, with a discretization of size $X$ and $U$ for the state and input spaces, respectively, the proposed approach reduces the time complexity of each iteration in the VI algorithm from $O(XU)$ to $O(X+U)$, by replacing the minimization operation in the primal domain with a simple addition in the conjugate domain.
| null |
Adaptive Risk Minimization: Learning to Adapt to Domain Shift
|
https://papers.nips.cc/paper_files/paper/2021/hash/c705112d1ec18b97acac7e2d63973424-Abstract.html
|
Marvin Zhang, Henrik Marklund, Nikita Dhawan, Abhishek Gupta, Sergey Levine, Chelsea Finn
|
https://papers.nips.cc/paper_files/paper/2021/hash/c705112d1ec18b97acac7e2d63973424-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13435-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c705112d1ec18b97acac7e2d63973424-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=-zgb2v8vV_w
|
https://papers.nips.cc/paper_files/paper/2021/file/c705112d1ec18b97acac7e2d63973424-Supplemental.pdf
|
A fundamental assumption of most machine learning algorithms is that the training and test data are drawn from the same underlying distribution. However, this assumption is violated in almost all practical applications: machine learning systems are regularly tested under distribution shift, due to changing temporal correlations, atypical end users, or other factors. In this work, we consider the problem setting of domain generalization, where the training data are structured into domains and there may be multiple test time shifts, corresponding to new domains or domain distributions. Most prior methods aim to learn a single robust model or invariant feature space that performs well on all domains. In contrast, we aim to learn models that adapt at test time to domain shift using unlabeled test points. Our primary contribution is to introduce the framework of adaptive risk minimization (ARM), in which models are directly optimized for effective adaptation to shift by learning to adapt on the training domains. Compared to prior methods for robustness, invariance, and adaptation, ARM methods provide performance gains of 1-4% test accuracy on a number of image classification problems exhibiting domain shift.
| null |
Learning State Representations from Random Deep Action-conditional Predictions
|
https://papers.nips.cc/paper_files/paper/2021/hash/c71df24045cfddab4a963d3ac9bdc9a3-Abstract.html
|
Zeyu Zheng, Vivek Veeriah, Risto Vuorio, Richard L Lewis, Satinder Singh
|
https://papers.nips.cc/paper_files/paper/2021/hash/c71df24045cfddab4a963d3ac9bdc9a3-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13436-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c71df24045cfddab4a963d3ac9bdc9a3-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=nJqCQUzpvS
|
https://papers.nips.cc/paper_files/paper/2021/file/c71df24045cfddab4a963d3ac9bdc9a3-Supplemental.zip
|
Our main contribution in this work is an empirical finding that random General Value Functions (GVFs), i.e., deep action-conditional predictions---random both in what feature of observations they predict as well as in the sequence of actions the predictions are conditioned upon---form good auxiliary tasks for reinforcement learning (RL) problems. In particular, we show that random deep action-conditional predictions when used as auxiliary tasks yield state representations that produce control performance competitive with state-of-the-art hand-crafted auxiliary tasks like value prediction, pixel control, and CURL in both Atari and DeepMind Lab tasks. In another set of experiments we stop the gradients from the RL part of the network to the state representation learning part of the network and show, perhaps surprisingly, that the auxiliary tasks alone are sufficient to learn state representations good enough to outperform an end-to-end trained actor-critic baseline. We opensourced our code at https://github.com/Hwhitetooth/random_gvfs.
| null |
Mixability made efficient: Fast online multiclass logistic regression
|
https://papers.nips.cc/paper_files/paper/2021/hash/c74214a3877c4d8297ac96217d5189b7-Abstract.html
|
Rémi Jézéquel, Pierre Gaillard, Alessandro Rudi
|
https://papers.nips.cc/paper_files/paper/2021/hash/c74214a3877c4d8297ac96217d5189b7-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13437-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c74214a3877c4d8297ac96217d5189b7-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=ejmqyWW0MK6
|
https://papers.nips.cc/paper_files/paper/2021/file/c74214a3877c4d8297ac96217d5189b7-Supplemental.pdf
|
Mixability has been shown to be a powerful tool to obtain algorithms with optimal regret. However, the resulting methods often suffer from high computational complexity which has reduced their practical applicability. For example, in the case of multiclass logistic regression, the aggregating forecaster (see Foster et al. 2018) achieves a regret of $O(\log(Bn))$ whereas Online Newton Step achieves $O(e^B\log(n))$ obtaining a double exponential gain in $B$ (a bound on the norm of comparative functions). However, this high statistical performance is at the price of a prohibitive computational complexity $O(n^{37})$.In this paper, we use quadratic surrogates to make aggregating forecasters more efficient. We show that the resulting algorithm has still high statistical performance for a large class of losses. In particular, we derive an algorithm for multiclass regression with a regret bounded by $O(B\log(n))$ and computational complexity of only $O(n^4)$.
| null |
Tracking People with 3D Representations
|
https://papers.nips.cc/paper_files/paper/2021/hash/c74c4bf0dad9cbae3d80faa054b7d8ca-Abstract.html
|
Jathushan Rajasegaran, Georgios Pavlakos, Angjoo Kanazawa, Jitendra Malik
|
https://papers.nips.cc/paper_files/paper/2021/hash/c74c4bf0dad9cbae3d80faa054b7d8ca-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13438-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c74c4bf0dad9cbae3d80faa054b7d8ca-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=NP-9Ppxdca
|
https://papers.nips.cc/paper_files/paper/2021/file/c74c4bf0dad9cbae3d80faa054b7d8ca-Supplemental.pdf
|
We present a novel approach for tracking multiple people in video. Unlike past approaches which employ 2D representations, we focus on using 3D representations of people, located in three-dimensional space. To this end, we develop a method, Human Mesh and Appearance Recovery (HMAR) which in addition to extracting the 3D geometry of the person as a SMPL mesh, also extracts appearance as a texture map on the triangles of the mesh. This serves as a 3D representation for appearance that is robust to viewpoint and pose changes. Given a video clip, we first detect bounding boxes corresponding to people, and for each one, we extract 3D appearance, pose, and location information using HMAR. These embedding vectors are then sent to a transformer, which performs spatio-temporal aggregation of the representations over the duration of the sequence. The similarity of the resulting representations is used to solve for associations that assigns each person to a tracklet. We evaluate our approach on the Posetrack, MuPoTs and AVA datasets. We find that 3D representations are more effective than 2D representations for tracking in these settings, and we obtain state-of-the-art performance. Code and results are available at: https://brjathu.github.io/T3DP.
| null |
Off-Policy Risk Assessment in Contextual Bandits
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7502c55f8db540625b59d9a42638520-Abstract.html
|
Audrey Huang, Liu Leqi, Zachary Lipton, Kamyar Azizzadenesheli
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7502c55f8db540625b59d9a42638520-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13439-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c7502c55f8db540625b59d9a42638520-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=u9RvlvaBC7
|
https://papers.nips.cc/paper_files/paper/2021/file/c7502c55f8db540625b59d9a42638520-Supplemental.pdf
|
Even when unable to run experiments, practitioners can evaluate prospective policies, using previously logged data. However, while the bandits literature has adopted a diverse set of objectives, most research on off-policy evaluation to date focuses on the expected reward. In this paper, we introduce Lipschitz risk functionals, a broad class of objectives that subsumes conditional value-at-risk (CVaR), variance, mean-variance, many distorted risks, and CPT risks, among others. We propose Off-Policy Risk Assessment (OPRA), a framework that first estimates a target policy's CDF and then generates plugin estimates for any collection of Lipschitz risks, providing finite sample guarantees that hold simultaneously over the entire class. We instantiate OPRA with both importance sampling and doubly robust estimators. Our primary theoretical contributions are (i) the first uniform concentration inequalities for both CDF estimators in contextual bandits and (ii) error bounds on our Lipschitz risk estimates, which all converge at a rate of $O(1/\sqrt{n})$.
| null |
Adaptive Denoising via GainTuning
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7558e9d1f956b016d1fdba7ea132378-Abstract.html
|
Sreyas Mohan, Joshua L Vincent, Ramon Manzorro, Peter Crozier, Carlos Fernandez-Granda, Eero Simoncelli
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7558e9d1f956b016d1fdba7ea132378-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13440-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c7558e9d1f956b016d1fdba7ea132378-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=0BHU7WvZ29
|
https://papers.nips.cc/paper_files/paper/2021/file/c7558e9d1f956b016d1fdba7ea132378-Supplemental.pdf
|
Deep convolutional neural networks (CNNs) for image denoising are typically trained on large datasets. These models achieve the current state of the art, but they do not generalize well to data that deviate from the training distribution. Recent work has shown that it is possible to train denoisers on a single noisy image. These models adapt to the features of the test image, but their performance is limited by the small amount of information used to train them. Here we propose "GainTuning'', a methodology by which CNN models pre-trained on large datasets can be adaptively and selectively adjusted for individual test images. To avoid overfitting, GainTuning optimizes a single multiplicative scaling parameter (the “Gain”) of each channel in the convolutional layers of the CNN. We show that GainTuning improves state-of-the-art CNNs on standard image-denoising benchmarks, boosting their denoising performance on nearly every image in a held-out test set. These adaptive improvements are even more substantial for test images differing systematically from the training data, either in noise level or image type. We illustrate the potential of adaptive GainTuning in a scientific application to transmission-electron-microscope images, using a CNN that is pre-trained on synthetic data. In contrast to the existing methodology, GainTuning is able to faithfully reconstruct the structure of catalytic nanoparticles from these data at extremely low signal-to-noise ratios.
| null |
Optimal Sketching for Trace Estimation
|
https://papers.nips.cc/paper_files/paper/2021/hash/c77bfda61a0204d445185053e6a9a8fe-Abstract.html
|
Shuli Jiang, Hai Pham, David Woodruff, Richard Zhang
|
https://papers.nips.cc/paper_files/paper/2021/hash/c77bfda61a0204d445185053e6a9a8fe-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13441-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c77bfda61a0204d445185053e6a9a8fe-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=rxud5HYKX55
|
https://papers.nips.cc/paper_files/paper/2021/file/c77bfda61a0204d445185053e6a9a8fe-Supplemental.pdf
|
Matrix trace estimation is ubiquitous in machine learning applications and has traditionally relied on Hutchinson's method, which requires $O(\log(1/\delta)/\epsilon^2)$ matrix-vector product queries to achieve a $(1 \pm \epsilon)$-multiplicative approximation to $\text{trace}(A)$ with failure probability $\delta$ on positive-semidefinite input matrices $A$. Recently, the Hutch++ algorithm was proposed, which reduces the number of matrix-vector queries from $O(1/\epsilon^2)$ to the optimal $O(1/\epsilon)$, and the algorithm succeeds with constant probability. However, in the high probability setting, the non-adaptive Hutch++ algorithm suffers an extra $O(\sqrt{\log(1/\delta)})$ multiplicative factor in its query complexity. Non-adaptive methods are important, as they correspond to sketching algorithms, which are mergeable, highly parallelizable, and provide low-memory streaming algorithms as well as low-communication distributed protocols. In this work, we close the gap between non-adaptive and adaptive algorithms, showing that even non-adaptive algorithms can achieve $O(\sqrt{\log(1/\delta)}/\epsilon + \log(1/\delta))$ matrix-vector products. In addition, we prove matching lower bounds demonstrating that, up to a $\log \log(1/\delta)$ factor, no further improvement in the dependence on $\delta$ or $\epsilon$ is possible by any non-adaptive algorithm. Finally, our experiments demonstrate the superior performance of our sketch over the adaptive Hutch++ algorithm, which is less parallelizable, as well as over the non-adaptive Hutchinson's method.
| null |
Estimating Multi-cause Treatment Effects via Single-cause Perturbation
|
https://papers.nips.cc/paper_files/paper/2021/hash/c793b3be8f18731f2a4c627fb3c6c63d-Abstract.html
|
Zhaozhi Qian, Alicia Curth, Mihaela van der Schaar
|
https://papers.nips.cc/paper_files/paper/2021/hash/c793b3be8f18731f2a4c627fb3c6c63d-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13442-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c793b3be8f18731f2a4c627fb3c6c63d-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=oz3t1BrfNO
|
https://papers.nips.cc/paper_files/paper/2021/file/c793b3be8f18731f2a4c627fb3c6c63d-Supplemental.pdf
|
Most existing methods for conditional average treatment effect estimation are designed to estimate the effect of a single cause - only one variable can be intervened on at one time. However, many applications involve simultaneous intervention on multiple variables, which leads to multi-cause treatment effect problems. The multi-cause problem is challenging because one needs to overcome the confounding bias for a large number of treatment groups, each with a different cause combination. The combinatorial nature of the problem also leads to severe data scarcity - we only observe one factual outcome out of many potential outcomes. In this work, we propose Single-cause Perturbation (SCP), a novel two-step procedure to estimate the multi-cause treatment effect. SCP starts by augmenting the observational dataset with the estimated potential outcomes under single-cause interventions. It then performs covariate adjustment on the augmented dataset to obtain the estimator. SCP is agnostic to the exact choice of algorithm in either step. We show formally that the procedure is valid under standard assumptions in causal inference. We demonstrate the performance gain of SCP on extensive synthetic and semi-synthetic experiments.
| null |
Be Confident! Towards Trustworthy Graph Neural Networks via Confidence Calibration
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7a9f13a6c0940277d46706c7ca32601-Abstract.html
|
Xiao Wang, Hongrui Liu, Chuan Shi, Cheng Yang
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7a9f13a6c0940277d46706c7ca32601-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13443-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c7a9f13a6c0940277d46706c7ca32601-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=9c-IsSptbmA
|
https://papers.nips.cc/paper_files/paper/2021/file/c7a9f13a6c0940277d46706c7ca32601-Supplemental.pdf
|
Despite Graph Neural Networks (GNNs) have achieved remarkable accuracy, whether the results are trustworthy is still unexplored. Previous studies suggest that many modern neural networks are over-confident on the predictions, however, surprisingly, we discover that GNNs are primarily in the opposite direction, i.e., GNNs are under-confident. Therefore, the confidence calibration for GNNs is highly desired. In this paper, we propose a novel trustworthy GNN model by designing a topology-aware post-hoc calibration function. Specifically, we first verify that the confidence distribution in a graph has homophily property, and this finding inspires us to design a calibration GNN model (CaGCN) to learn the calibration function. CaGCN is able to obtain a unique transformation from logits of GNNs to the calibrated confidence for each node, meanwhile, such transformation is able to preserve the order between classes, satisfying the accuracy-preserving property. Moreover, we apply the calibration GNN to self-training framework, showing that more trustworthy pseudo labels can be obtained with the calibrated confidence and further improve the performance. Extensive experiments demonstrate the effectiveness of our proposed model in terms of both calibration and accuracy.
| null |
Learning Riemannian metric for disease progression modeling
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7b90b0fc23725f299b47c5224e6ec0d-Abstract.html
|
Samuel Gruffaz, Pierre-Emmanuel Poulet, Etienne Maheux, Bruno Jedynak, Stanley DURRLEMAN
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7b90b0fc23725f299b47c5224e6ec0d-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13444-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c7b90b0fc23725f299b47c5224e6ec0d-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=qZpOqPbwhy
|
https://papers.nips.cc/paper_files/paper/2021/file/c7b90b0fc23725f299b47c5224e6ec0d-Supplemental.pdf
|
Linear mixed-effect models provide a natural baseline for estimating disease progression using longitudinal data. They provide interpretable models at the cost of modeling assumptions on the progression profiles and their variability across subjects. A significant improvement is to embed the data in a Riemannian manifold and learn patient-specific trajectories distributed around a central geodesic. A few interpretable parameters characterize subject trajectories at the cost of a prior choice of the metric, which determines the shape of the trajectories. We extend this approach by learning the metric from the data allowing more flexibility while keeping the interpretability. Specifically, we learn the metric as the push-forward of the Euclidean metric by a diffeomorphism. This diffeomorphism is estimated iteratively as the composition of radial basis functions belonging to a reproducible kernel Hilbert space. The metric update allows us to improve the forecasting of imaging and clinical biomarkers in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort. Our results compare favorably to the 56 methods benchmarked in the TADPOLE challenge.
| null |
Bias and variance of the Bayesian-mean decoder
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7c3e78e3c9d26cc1158a8735d548eaa-Abstract.html
|
Arthur Prat-Carrabin, Michael Woodford
|
https://papers.nips.cc/paper_files/paper/2021/hash/c7c3e78e3c9d26cc1158a8735d548eaa-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13445-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c7c3e78e3c9d26cc1158a8735d548eaa-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=7rYDxRb1eSa
|
https://papers.nips.cc/paper_files/paper/2021/file/c7c3e78e3c9d26cc1158a8735d548eaa-Supplemental.pdf
|
Perception, in theoretical neuroscience, has been modeled as the encoding of external stimuli into internal signals, which are then decoded. The Bayesian mean is an important decoder, as it is optimal for purposes of both estimation and discrimination. We present widely-applicable approximations to the bias and to the variance of the Bayesian mean, obtained under the minimal and biologically-relevant assumption that the encoding results from a series of independent, though not necessarily identically-distributed, signals. Simulations substantiate the accuracy of our approximations in the small-noise regime. The bias of the Bayesian mean comprises two components: one driven by the prior, and one driven by the precision of the encoding. If the encoding is 'efficient', the two components have opposite effects; their relative strengths are determined by the objective that the encoding optimizes. The experimental literature on perception reports both 'Bayesian' biases directed towards prior expectations, and opposite, 'anti-Bayesian' biases. We show that different tasks are indeed predicted to yield such contradictory biases, under a consistently-optimal encoding-decoding model. Moreover, we recover Wei and Stocker's "law of human perception", a relation between the bias of the Bayesian mean and the derivative of its variance, and show how the coefficient of proportionality in this law depends on the task at hand. Our results provide a parsimonious theory of optimal perception under constraints, in which encoding and decoding are adapted both to the prior and to the task faced by the observer.
| null |
MIRACLE: Causally-Aware Imputation via Learning Missing Data Mechanisms
|
https://papers.nips.cc/paper_files/paper/2021/hash/c80bcf42c220b8f5c41f85344242f1b0-Abstract.html
|
Trent Kyono, Yao Zhang, Alexis Bellot, Mihaela van der Schaar
|
https://papers.nips.cc/paper_files/paper/2021/hash/c80bcf42c220b8f5c41f85344242f1b0-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13446-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c80bcf42c220b8f5c41f85344242f1b0-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=GzeqcAUFGl0
|
https://papers.nips.cc/paper_files/paper/2021/file/c80bcf42c220b8f5c41f85344242f1b0-Supplemental.pdf
|
Missing data is an important problem in machine learning practice. Starting from the premise that imputation methods should preserve the causal structure of the data, we develop a regularization scheme that encourages any baseline imputation method to be causally consistent with the underlying data generating mechanism. Our proposal is a causally-aware imputation algorithm (MIRACLE). MIRACLE iteratively refines the imputation of a baseline by simultaneously modeling the missingness generating mechanism, encouraging imputation to be consistent with the causal structure of the data. We conduct extensive experiments on synthetic and a variety of publicly available datasets to show that MIRACLE is able to consistently improve imputation over a variety of benchmark methods across all three missingness scenarios: at random, completely at random, and not at random.
| null |
Efficient Training of Visual Transformers with Small Datasets
|
https://papers.nips.cc/paper_files/paper/2021/hash/c81e155d85dae5430a8cee6f2242e82c-Abstract.html
|
Yahui Liu, Enver Sangineto, Wei Bi, Nicu Sebe, Bruno Lepri, Marco Nadai
|
https://papers.nips.cc/paper_files/paper/2021/hash/c81e155d85dae5430a8cee6f2242e82c-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13447-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c81e155d85dae5430a8cee6f2242e82c-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=SCN8UaetXx
|
https://papers.nips.cc/paper_files/paper/2021/file/c81e155d85dae5430a8cee6f2242e82c-Supplemental.pdf
|
Visual Transformers (VTs) are emerging as an architectural paradigm alternative to Convolutional networks (CNNs). Differently from CNNs, VTs can capture global relations between image elements and they potentially have a larger representation capacity. However, the lack of the typical convolutional inductive bias makes these models more data hungry than common CNNs. In fact, some local properties of the visual domain which are embedded in the CNN architectural design, in VTs should be learned from samples. In this paper, we empirically analyse different VTs, comparing their robustness in a small training set regime, and we show that, despite having a comparable accuracy when trained on ImageNet, their performance on smaller datasets can be largely different. Moreover, we propose an auxiliary self-supervised task which can extract additional information from images with only a negligible computational overhead. This task encourages the VTs to learn spatial relations within an image and makes the VT training much more robust when training data is scarce. Our task is used jointly with the standard (supervised) training and it does not depend on specific architectural choices, thus it can be easily plugged in the existing VTs. Using an extensive evaluation with different VTs and datasets, we show that our method can improve (sometimes dramatically) the final accuracy of the VTs. Our code is available at: https://github.com/yhlleo/VTs-Drloc.
| null |
Small random initialization is akin to spectral learning: Optimization and generalization guarantees for overparameterized low-rank matrix reconstruction
|
https://papers.nips.cc/paper_files/paper/2021/hash/c82836ed448c41094025b4a872c5341e-Abstract.html
|
Dominik Stöger, Mahdi Soltanolkotabi
|
https://papers.nips.cc/paper_files/paper/2021/hash/c82836ed448c41094025b4a872c5341e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13448-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c82836ed448c41094025b4a872c5341e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=rsRq--gsiE
|
https://papers.nips.cc/paper_files/paper/2021/file/c82836ed448c41094025b4a872c5341e-Supplemental.pdf
|
Recently there has been significant theoretical progress on understanding the convergence and generalization of gradient-based methods on nonconvex losses with overparameterized models. Nevertheless, many aspects of optimization and generalization and in particular the critical role of small random initialization are not fully understood. In this paper, we take a step towards demystifying this role by proving that small random initialization followed by a few iterations of gradient descent behaves akin to popular spectral methods. We also show that this implicit spectral bias from small random initialization, which is provably more prominent for overparameterized models, also puts the gradient descent iterations on a particular trajectory towards solutions that are not only globally optimal but also generalize well. Concretely, we focus on the problem of reconstructing a low-rank matrix from a few measurements via a natural nonconvex formulation. In this setting, we show that the trajectory of the gradient descent iterations from small random initialization can be approximately decomposed into three phases: (I) a spectral or alignment phase where we show that that the iterates have an implicit spectral bias akin to spectral initialization allowing us to show that at the end of this phase the column space of the iterates and the underlying low-rank matrix are sufficiently aligned, (II) a saddle avoidance/refinement phase where we show that the trajectory of the gradient iterates moves away from certain degenerate saddle points, and (III) a local refinement phase where we show that after avoiding the saddles the iterates converge quickly to the underlying low-rank matrix. Underlying our analysis are insights for the analysis of overparameterized nonconvex optimization schemes that may have implications for computational problems beyond low-rank reconstruction.
| null |
Efficient Combination of Rematerialization and Offloading for Training DNNs
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8461bf13fca8a2b9912ab2eb1668e4b-Abstract.html
|
Olivier Beaumont, Lionel Eyraud-Dubois, Alena Shilova
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8461bf13fca8a2b9912ab2eb1668e4b-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13449-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c8461bf13fca8a2b9912ab2eb1668e4b-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=BFYlnDtJSqW
|
https://papers.nips.cc/paper_files/paper/2021/file/c8461bf13fca8a2b9912ab2eb1668e4b-Supplemental.zip
|
Rematerialization and offloading are two well known strategies to save memory during the training phase of deep neural networks, allowing data scientists to consider larger models, batch sizes or higher resolution data. Rematerialization trades memory for computation time, whereas Offloading trades memory for data movements. As these two resources are independent, it is appealing to consider the simultaneous combination of both strategies to save even more memory. We precisely model the costs and constraints corresponding to Deep Learning frameworks such as PyTorch or Tensorflow, we propose optimal algorithms to find a valid sequence of memory-constrained operations and finally, we evaluate the performance of proposed algorithms on realistic networks and computation platforms. Our experiments show that the possibility to offload can remove one third of the overhead of rematerialization, and that together they can reduce the memory used for activations by a factor 4 to 6, with an overhead below 20%.
| null |
Particle Cloud Generation with Message Passing Generative Adversarial Networks
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8512d142a2d849725f31a9a7a361ab9-Abstract.html
|
Raghav Kansal, Javier Duarte, Hao Su, Breno Orzari, Thiago Tomei, Maurizio Pierini, Mary Touranakou, jean-roch vlimant, Dimitrios Gunopulos
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8512d142a2d849725f31a9a7a361ab9-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13450-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c8512d142a2d849725f31a9a7a361ab9-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=iorEu783qJ5
|
https://papers.nips.cc/paper_files/paper/2021/file/c8512d142a2d849725f31a9a7a361ab9-Supplemental.pdf
|
In high energy physics (HEP), jets are collections of correlated particles produced ubiquitously in particle collisions such as those at the CERN Large Hadron Collider (LHC). Machine learning (ML)-based generative models, such as generative adversarial networks (GANs), have the potential to significantly accelerate LHC jet simulations. However, despite jets having a natural representation as a set of particles in momentum-space, a.k.a. a particle cloud, there exist no generative models applied to such a dataset. In this work, we introduce a new particle cloud dataset (JetNet), and apply to it existing point cloud GANs. Results are evaluated using (1) 1-Wasserstein distances between high- and low-level feature distributions, (2) a newly developed Fréchet ParticleNet Distance, and (3) the coverage and (4) minimum matching distance metrics. Existing GANs are found to be inadequate for physics applications, hence we develop a new message passing GAN (MPGAN), which outperforms existing point cloud GANs on virtually every metric and shows promise for use in HEP. We propose JetNet as a novel point-cloud-style dataset for the ML community to experiment with, and set MPGAN as a benchmark to improve upon for future generative models. Additionally, to facilitate research and improve accessibility and reproducibility in this area, we release the open-source JetNet Python package with interfaces for particle cloud datasets, implementations for evaluation and loss metrics, and more tools for ML in HEP development.
| null |
CoFiNet: Reliable Coarse-to-fine Correspondences for Robust PointCloud Registration
|
https://papers.nips.cc/paper_files/paper/2021/hash/c85b2ea9a678e74fdc8bafe5d0707c31-Abstract.html
|
Hao Yu, Fu Li, Mahdi Saleh, Benjamin Busam, Slobodan Ilic
|
https://papers.nips.cc/paper_files/paper/2021/hash/c85b2ea9a678e74fdc8bafe5d0707c31-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13451-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c85b2ea9a678e74fdc8bafe5d0707c31-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=HhUmPH22Vpn
|
https://papers.nips.cc/paper_files/paper/2021/file/c85b2ea9a678e74fdc8bafe5d0707c31-Supplemental.pdf
|
We study the problem of extracting correspondences between a pair of point clouds for registration. For correspondence retrieval, existing works benefit from matching sparse keypoints detected from dense points but usually struggle to guarantee their repeatability. To address this issue, we present CoFiNet - Coarse-to-Fine Network which extracts hierarchical correspondences from coarse to fine without keypoint detection. On a coarse scale and guided by a weighting scheme, our model firstly learns to match down-sampled nodes whose vicinity points share more overlap, which significantly shrinks the search space of a consecutive stage. On a finer scale, node proposals are consecutively expanded to patches that consist of groups of points together with associated descriptors. Point correspondences are then refined from the overlap areas of corresponding patches, by a density-adaptive matching module capable to deal with varying point density. Extensive evaluation of CoFiNet on both indoor and outdoor standard benchmarks shows our superiority over existing methods. Especially on 3DLoMatch where point clouds share less overlap, CoFiNet significantly outperforms state-of-the-art approaches by at least 5% on Registration Recall, with at most two-third of their parameters.
| null |
Partial success in closing the gap between human and machine vision
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8877cff22082a16395a57e97232bb6f-Abstract.html
|
Robert Geirhos, Kantharaju Narayanappa, Benjamin Mitzkus, Tizian Thieringer, Matthias Bethge, Felix A. Wichmann, Wieland Brendel
|
https://papers.nips.cc/paper_files/paper/2021/hash/c8877cff22082a16395a57e97232bb6f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13452-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c8877cff22082a16395a57e97232bb6f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=QkljT4mrfs
|
https://papers.nips.cc/paper_files/paper/2021/file/c8877cff22082a16395a57e97232bb6f-Supplemental.pdf
|
A few years ago, the first CNN surpassed human performance on ImageNet. However, it soon became clear that machines lack robustness on more challenging test cases, a major obstacle towards deploying machines "in the wild" and towards obtaining better computational models of human visual perception. Here we ask: Are we making progress in closing the gap between human and machine vision? To answer this question, we tested human observers on a broad range of out-of-distribution (OOD) datasets, recording 85,120 psychophysical trials across 90 participants. We then investigated a range of promising machine learning developments that crucially deviate from standard supervised CNNs along three axes: objective function (self-supervised, adversarially trained, CLIP language-image training), architecture (e.g. vision transformers), and dataset size (ranging from 1M to 1B).Our findings are threefold. (1.) The longstanding distortion robustness gap between humans and CNNs is closing, with the best models now exceeding human feedforward performance on most of the investigated OOD datasets. (2.) There is still a substantial image-level consistency gap, meaning that humans make different errors than models. In contrast, most models systematically agree in their categorisation errors, even substantially different ones like contrastive self-supervised vs. standard supervised models. (3.) In many cases, human-to-model consistency improves when training dataset size is increased by one to three orders of magnitude. Our results give reason for cautious optimism: While there is still much room for improvement, the behavioural difference between human and machine vision is narrowing. In order to measure future progress, 17 OOD datasets with image-level human behavioural data and evaluation code are provided as a toolbox and benchmark at: https://github.com/bethgelab/model-vs-human/
| null |
LLC: Accurate, Multi-purpose Learnt Low-dimensional Binary Codes
|
https://papers.nips.cc/paper_files/paper/2021/hash/c88d8d0a6097754525e02c2246d8d27f-Abstract.html
|
Aditya Kusupati, Matthew Wallingford, Vivek Ramanujan, Raghav Somani, Jae Sung Park, Krishna Pillutla, Prateek Jain, Sham Kakade, Ali Farhadi
|
https://papers.nips.cc/paper_files/paper/2021/hash/c88d8d0a6097754525e02c2246d8d27f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13453-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c88d8d0a6097754525e02c2246d8d27f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=oiq92o1EFg1
|
https://papers.nips.cc/paper_files/paper/2021/file/c88d8d0a6097754525e02c2246d8d27f-Supplemental.pdf
|
Learning binary representations of instances and classes is a classical problem with several high potential applications. In modern settings, the compression of high-dimensional neural representations to low-dimensional binary codes is a challenging task and often require large bit-codes to be accurate. In this work, we propose a novel method for $\textbf{L}$earning $\textbf{L}$ow-dimensional binary $\textbf{C}$odes $(\textbf{LLC})$ for instances as well as classes. Our method does ${\textit{not}}$ require any side-information, like annotated attributes or label meta-data, and learns extremely low-dimensional binary codes ($\approx 20$ bits for ImageNet-1K). The learnt codes are super-efficient while still ensuring $\textit{nearly optimal}$ classification accuracy for ResNet50 on ImageNet-1K. We demonstrate that the learnt codes capture intrinsically important features in the data, by discovering an intuitive taxonomy over classes. We further quantitatively measure the quality of our codes by applying it to the efficient image retrieval as well as out-of-distribution (OOD) detection problems. For ImageNet-100 retrieval problem, our learnt binary codes outperform $16$ bit HashNet using only $10$ bits and also are as accurate as $10$ dimensional real representations. Finally, our learnt binary codes can perform OOD detection, out-of-the-box, as accurately as a baseline that needs $\approx3000$ samples to tune its threshold, while we require ${\textit{none}}$. Code is open-sourced at https://github.com/RAIVNLab/LLC.
| null |
Analytic Insights into Structure and Rank of Neural Network Hessian Maps
|
https://papers.nips.cc/paper_files/paper/2021/hash/c900ced7451da79502d29aa37ebb7b60-Abstract.html
|
Sidak Pal Singh, Gregor Bachmann, Thomas Hofmann
|
https://papers.nips.cc/paper_files/paper/2021/hash/c900ced7451da79502d29aa37ebb7b60-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13454-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c900ced7451da79502d29aa37ebb7b60-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=otDgw7LM7Nn
|
https://papers.nips.cc/paper_files/paper/2021/file/c900ced7451da79502d29aa37ebb7b60-Supplemental.pdf
|
The Hessian of a neural network captures parameter interactions through second-order derivatives of the loss. It is a fundamental object of study, closely tied to various problems in deep learning, including model design, optimization, and generalization. Most prior work has been empirical, typically focusing on low-rank approximations and heuristics that are blind to the network structure. In contrast, we develop theoretical tools to analyze the range of the Hessian map, which provide us with a precise understanding of its rank deficiency and the structural reasons behind it. This yields exact formulas and tight upper bounds for the Hessian rank of deep linear networks --- allowing for an elegant interpretation in terms of rank deficiency. Moreover, we demonstrate that our bounds remain faithful as an estimate of the numerical Hessian rank, for a larger class of models such as rectified and hyperbolic tangent networks. Further, we also investigate the implications of model architecture (e.g.~width, depth, bias) on the rank deficiency. Overall, our work provides novel insights into the source and extent of redundancy in overparameterized neural networks.
| null |
Well-tuned Simple Nets Excel on Tabular Datasets
|
https://papers.nips.cc/paper_files/paper/2021/hash/c902b497eb972281fb5b4e206db38ee6-Abstract.html
|
Arlind Kadra, Marius Lindauer, Frank Hutter, Josif Grabocka
|
https://papers.nips.cc/paper_files/paper/2021/hash/c902b497eb972281fb5b4e206db38ee6-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13455-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c902b497eb972281fb5b4e206db38ee6-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=d3k38LTDCyO
|
https://papers.nips.cc/paper_files/paper/2021/file/c902b497eb972281fb5b4e206db38ee6-Supplemental.pdf
|
Tabular datasets are the last "unconquered castle" for deep learning, with traditional ML methods like Gradient-Boosted Decision Trees still performing strongly even against recent specialized neural architectures. In this paper, we hypothesize that the key to boosting the performance of neural networks lies in rethinking the joint and simultaneous application of a large set of modern regularization techniques. As a result, we propose regularizing plain Multilayer Perceptron (MLP) networks by searching for the optimal combination/cocktail of 13 regularization techniques for each dataset using a joint optimization over the decision on which regularizers to apply and their subsidiary hyperparameters. We empirically assess the impact of these regularization cocktails for MLPs in a large-scale empirical study comprising 40 tabular datasets and demonstrate that (i) well-regularized plain MLPs significantly outperform recent state-of-the-art specialized neural network architectures, and (ii) they even outperform strong traditional ML methods, such as XGBoost.
| null |
POODLE: Improving Few-shot Learning via Penalizing Out-of-Distribution Samples
|
https://papers.nips.cc/paper_files/paper/2021/hash/c91591a8d461c2869b9f535ded3e213e-Abstract.html
|
Duong Le, Khoi Duc Nguyen, Khoi Nguyen, Quoc-Huy Tran, Rang Nguyen, Binh-Son Hua
|
https://papers.nips.cc/paper_files/paper/2021/hash/c91591a8d461c2869b9f535ded3e213e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13456-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c91591a8d461c2869b9f535ded3e213e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=sfzseGUqFrd
|
https://papers.nips.cc/paper_files/paper/2021/file/c91591a8d461c2869b9f535ded3e213e-Supplemental.pdf
|
In this work, we propose to use out-of-distribution samples, i.e., unlabeled samples coming from outside the target classes, to improve few-shot learning. Specifically, we exploit the easily available out-of-distribution samples to drive the classifier to avoid irrelevant features by maximizing the distance from prototypes to out-of-distribution samples while minimizing that of in-distribution samples (i.e., support, query data). Our approach is simple to implement, agnostic to feature extractors, lightweight without any additional cost for pre-training, and applicable to both inductive and transductive settings. Extensive experiments on various standard benchmarks demonstrate that the proposed method consistently improves the performance of pretrained networks with different architectures.
| null |
Combinatorial Pure Exploration with Bottleneck Reward Function
|
https://papers.nips.cc/paper_files/paper/2021/hash/c92a10324374fac681719d63979d00fe-Abstract.html
|
Yihan Du, Yuko Kuroki, Wei Chen
|
https://papers.nips.cc/paper_files/paper/2021/hash/c92a10324374fac681719d63979d00fe-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13457-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c92a10324374fac681719d63979d00fe-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=mVt55ZQqfTl
|
https://papers.nips.cc/paper_files/paper/2021/file/c92a10324374fac681719d63979d00fe-Supplemental.pdf
|
In this paper, we study the Combinatorial Pure Exploration problem with the Bottleneck reward function (CPE-B) under the fixed-confidence (FC) and fixed-budget (FB) settings.In CPE-B, given a set of base arms and a collection of subsets of base arms (super arms) following a certain combinatorial constraint, a learner sequentially plays a base arm and observes its random reward, with the objective of finding the optimal super arm with the maximum bottleneck value, defined as the minimum expected reward of the base arms contained in the super arm.CPE-B captures a variety of practical scenarios such as network routing in communication networks, and its unique challenges fall on how to utilize the bottleneck property to save samples and achieve the statistical optimality. None of the existing CPE studies (most of them assume linear rewards) can be adapted to solve such challenges, and thus we develop brand-new techniques to handle them.For the FC setting, we propose novel algorithms with optimal sample complexity for a broad family of instances and establish a matching lower bound to demonstrate the optimality (within a logarithmic factor).For the FB setting, we design an algorithm which achieves the state-of-the-art error probability guarantee and is the first to run efficiently on fixed-budget path instances, compared to existing CPE algorithms. Our experimental results on the top-$k$, path and matching instances validate the empirical superiority of the proposed algorithms over their baselines.
| null |
Densely connected normalizing flows
|
https://papers.nips.cc/paper_files/paper/2021/hash/c950cde9b3f83f41721788e3315a14a3-Abstract.html
|
Matej Grcić, Ivan Grubišić, Siniša Šegvić
|
https://papers.nips.cc/paper_files/paper/2021/hash/c950cde9b3f83f41721788e3315a14a3-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13458-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c950cde9b3f83f41721788e3315a14a3-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=JNSwviqJhS
|
https://papers.nips.cc/paper_files/paper/2021/file/c950cde9b3f83f41721788e3315a14a3-Supplemental.pdf
|
Normalizing flows are bijective mappings between inputs and latent representations with a fully factorized distribution. They are very attractive due to exact likelihood evaluation and efficient sampling. However, their effective capacity is often insufficient since the bijectivity constraint limits the model width. We address this issue by incrementally padding intermediate representations with noise. We precondition the noise in accordance with previous invertible units, which we describe as cross-unit coupling. Our invertible glow-like modules increase the model expressivity by fusing a densely connected block with Nyström self-attention. We refer to our architecture as DenseFlow since both cross-unit and intra-module couplings rely on dense connectivity. Experiments show significant improvements due to the proposed contributions and reveal state-of-the-art density estimation under moderate computing budgets.
| null |
Snowflake: Scaling GNNs to high-dimensional continuous control via parameter freezing
|
https://papers.nips.cc/paper_files/paper/2021/hash/c952ce98517ac529c60744ac28364b03-Abstract.html
|
Charles Blake, Vitaly Kurin, Maximilian Igl, Shimon Whiteson
|
https://papers.nips.cc/paper_files/paper/2021/hash/c952ce98517ac529c60744ac28364b03-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13459-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c952ce98517ac529c60744ac28364b03-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=REjT_c1Eejk
|
https://papers.nips.cc/paper_files/paper/2021/file/c952ce98517ac529c60744ac28364b03-Supplemental.pdf
|
Recent research has shown that graph neural networks (GNNs) can learn policies for locomotion control that are as effective as a typical multi-layer perceptron (MLP), with superior transfer and multi-task performance. However, results have so far been limited to training on small agents, with the performance of GNNs deteriorating rapidly as the number of sensors and actuators grows. A key motivation for the use of GNNs in the supervised learning setting is their applicability to large graphs, but this benefit has not yet been realised for locomotion control. We show that poor scaling in GNNs is a result of increasingly unstable policy updates, caused by overfitting in parts of the network during training. To combat this, we introduce Snowflake, a GNN training method for high-dimensional continuous control that freezes parameters in selected parts of the network. Snowflake significantly boosts the performance of GNNs for locomotion control on large agents, now matching the performance of MLPs while offering superior transfer properties.
| null |
Subgame solving without common knowledge
|
https://papers.nips.cc/paper_files/paper/2021/hash/c96c08f8bb7960e11a1239352a479053-Abstract.html
|
Brian Zhang, Tuomas Sandholm
|
https://papers.nips.cc/paper_files/paper/2021/hash/c96c08f8bb7960e11a1239352a479053-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13460-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c96c08f8bb7960e11a1239352a479053-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=StbpmmlJbH
|
https://papers.nips.cc/paper_files/paper/2021/file/c96c08f8bb7960e11a1239352a479053-Supplemental.pdf
|
In imperfect-information games, subgame solving is significantly more challenging than in perfect-information games, but in the last few years, such techniques have been developed. They were the key ingredient to the milestone of superhuman play in no-limit Texas hold'em poker. Current subgame-solving techniques analyze the entire common-knowledge closure of the player's current information set, that is, the smallest set of nodes within which it is common knowledge that the current node lies. While this is acceptable in games like poker where the common-knowledge closure is relatively small, many practical games have more complex information structure, which renders the common-knowledge closure impractically large to enumerate or even reasonably approximate. We introduce an approach that overcomes this obstacle, by instead working with only low-order knowledge. Our approach allows an agent, upon arriving at an infoset, to basically prune any node that is no longer reachable, thereby massively reducing the game tree size relative to the common-knowledge subgame. We prove that, as is, our approach can increase exploitability compared to the blueprint strategy. However, we develop three avenues by which safety can be guaranteed. First, safety is guaranteed if the results of subgame solves are incorporated back into the blueprint. Second, we provide a method where safety is achieved by limiting the infosets at which subgame solving is performed. Third, we prove that our approach, when applied at every infoset reached during play, achieves a weaker notion of equilibrium, which we coin affine equilibrium, and which may be of independent interest. We show that affine equilibria cannot be exploited by any Nash strategy of the opponent, so an opponent who wishes to exploit must open herself to counter-exploitation. Even without the safety-guaranteeing additions, experiments on medium-sized games show that our approach always reduced exploitability in practical games even when applied at every infoset, and a depth-limited version of it led to---to our knowledge---the first strong AI for the challenge problem dark chess.
| null |
Fair Algorithms for Multi-Agent Multi-Armed Bandits
|
https://papers.nips.cc/paper_files/paper/2021/hash/c96ebeee051996333b6d70b2da6191b0-Abstract.html
|
Safwan Hossain, Evi Micha, Nisarg Shah
|
https://papers.nips.cc/paper_files/paper/2021/hash/c96ebeee051996333b6d70b2da6191b0-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13461-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c96ebeee051996333b6d70b2da6191b0-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=AlD5WD2ANIQ
|
https://papers.nips.cc/paper_files/paper/2021/file/c96ebeee051996333b6d70b2da6191b0-Supplemental.pdf
|
We propose a multi-agent variant of the classical multi-armed bandit problem, in which there are $N$ agents and $K$ arms, and pulling an arm generates a (possibly different) stochastic reward for each agent. Unlike the classical multi-armed bandit problem, the goal is not to learn the "best arm"; indeed, each agent may perceive a different arm to be the best for her personally. Instead, we seek to learn a fair distribution over the arms. Drawing on a long line of research in economics and computer science, we use the Nash social welfare as our notion of fairness. We design multi-agent variants of three classic multi-armed bandit algorithms and show that they achieve sublinear regret, which is now measured in terms of the lost Nash social welfare. We also extend a classical lower bound, establishing the optimality of one of our algorithms.
| null |
VAST: Value Function Factorization with Variable Agent Sub-Teams
|
https://papers.nips.cc/paper_files/paper/2021/hash/c97e7a5153badb6576d8939469f58336-Abstract.html
|
Thomy Phan, Fabian Ritz, Lenz Belzner, Philipp Altmann, Thomas Gabor, Claudia Linnhoff-Popien
|
https://papers.nips.cc/paper_files/paper/2021/hash/c97e7a5153badb6576d8939469f58336-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13462-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c97e7a5153badb6576d8939469f58336-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=hyJKKIhfxxT
|
https://papers.nips.cc/paper_files/paper/2021/file/c97e7a5153badb6576d8939469f58336-Supplemental.pdf
|
Value function factorization (VFF) is a popular approach to cooperative multi-agent reinforcement learning in order to learn local value functions from global rewards. However, state-of-the-art VFF is limited to a handful of agents in most domains. We hypothesize that this is due to the flat factorization scheme, where the VFF operator becomes a performance bottleneck with an increasing number of agents. Therefore, we propose VFF with variable agent sub-teams (VAST). VAST approximates a factorization for sub-teams which can be defined in an arbitrary way and vary over time, e.g., to adapt to different situations. The sub-team values are then linearly decomposed for all sub-team members. Thus, VAST can learn on a more focused and compact input representation of the original VFF operator. We evaluate VAST in three multi-agent domains and show that VAST can significantly outperform state-of-the-art VFF, when the number of agents is sufficiently large.
| null |
On the Stochastic Stability of Deep Markov Models
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9dd73f5cb96486f5e1e0680e841a550-Abstract.html
|
Jan Drgona, Sayak Mukherjee, Jiaxin Zhang, Frank Liu, Mahantesh Halappanavar
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9dd73f5cb96486f5e1e0680e841a550-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13463-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c9dd73f5cb96486f5e1e0680e841a550-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=Z2ZWIvNeVUl
|
https://papers.nips.cc/paper_files/paper/2021/file/c9dd73f5cb96486f5e1e0680e841a550-Supplemental.pdf
|
Deep Markov models (DMM) are generative models which are scalable and expressive generalization of Markov models for representation, learning, and inference problems. However, the fundamental stochastic stability guarantees of such models have not been thoroughly investigated. In this paper, we present a novel stability analysis method and provide sufficient conditions of DMM's stochastic stability. The proposed stability analysis is based on the contraction of probabilistic maps modeled by deep neural networks. We make connections between the spectral properties of neural network's weights and different types of used activation function on the stability and overall dynamic behavior of DMMs with Gaussian distributions. Based on the theory, we propose a few practical methods for designing constrained DMMs with guaranteed stability. We empirically substantiate our theoretical results via intuitive numerical experiments using the proposed stability constraints.
| null |
Multiwavelet-based Operator Learning for Differential Equations
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9e5c2b59d98488fe1070e744041ea0e-Abstract.html
|
Gaurav Gupta, Xiongye Xiao, Paul Bogdan
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9e5c2b59d98488fe1070e744041ea0e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13464-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c9e5c2b59d98488fe1070e744041ea0e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=LZDiWaC9CGL
|
https://papers.nips.cc/paper_files/paper/2021/file/c9e5c2b59d98488fe1070e744041ea0e-Supplemental.pdf
|
The solution of a partial differential equation can be obtained by computing the inverse operator map between the input and the solution space. Towards this end, we introduce a $\textit{multiwavelet-based neural operator learning scheme}$ that compresses the associated operator's kernel using fine-grained wavelets. By explicitly embedding the inverse multiwavelet filters, we learn the projection of the kernel onto fixed multiwavelet polynomial bases. The projected kernel is trained at multiple scales derived from using repeated computation of multiwavelet transform. This allows learning the complex dependencies at various scales and results in a resolution-independent scheme. Compare to the prior works, we exploit the fundamental properties of the operator's kernel which enable numerically efficient representation. We perform experiments on the Korteweg-de Vries (KdV) equation, Burgers' equation, Darcy Flow, and Navier-Stokes equation. Compared with the existing neural operator approaches, our model shows significantly higher accuracy and achieves state-of-the-art in a range of datasets. For the time-varying equations, the proposed method exhibits a ($2X-10X$) improvement ($0.0018$ ($0.0033$) relative $L2$ error for Burgers' (KdV) equation). By learning the mappings between function spaces, the proposed method has the ability to find the solution of a high-resolution input after learning from lower-resolution data.
| null |
Intermediate Layers Matter in Momentum Contrastive Self Supervised Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9f06258da6455f5bf50c5b9260efeff-Abstract.html
|
Aakash Kaku, Sahana Upadhya, Narges Razavian
|
https://papers.nips.cc/paper_files/paper/2021/hash/c9f06258da6455f5bf50c5b9260efeff-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13465-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/c9f06258da6455f5bf50c5b9260efeff-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=M5j42PvY65V
|
https://papers.nips.cc/paper_files/paper/2021/file/c9f06258da6455f5bf50c5b9260efeff-Supplemental.pdf
|
We show that bringing intermediate layers' representations of two augmented versions of an image closer together in self-supervised learning helps to improve the momentum contrastive (MoCo) method. To this end, in addition to the contrastive loss, we minimize the mean squared error between the intermediate layer representations or make their cross-correlation matrix closer to an identity matrix. Both loss objectives either outperform standard MoCo, or achieve similar performances on three diverse medical imaging datasets: NIH-Chest Xrays, Breast Cancer Histopathology, and Diabetic Retinopathy. The gains of the improved MoCo are especially large in a low-labeled data regime (e.g. 1% labeled data) with an average gain of 5% across three datasets. We analyze the models trained using our novel approach via feature similarity analysis and layer-wise probing. Our analysis reveals that models trained via our approach have higher feature reuse compared to a standard MoCo and learn informative features earlier in the network. Finally, by comparing the output probability distribution of models fine-tuned on small versus large labeled data, we conclude that our proposed method of pre-training leads to lower Kolmogorov–Smirnov distance, as compared to a standard MoCo. This provides additional evidence that our proposed method learns more informative features in the pre-training phase which could be leveraged in a low-labeled data regime.
| null |
An Efficient Pessimistic-Optimistic Algorithm for Stochastic Linear Bandits with General Constraints
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca460332316d6da84b08b9bcf39b687b-Abstract.html
|
Xin Liu, Bin Li, Pengyi Shi, Lei Ying
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca460332316d6da84b08b9bcf39b687b-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13466-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca460332316d6da84b08b9bcf39b687b-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=luGpyKCzOPI
| null |
This paper considers stochastic linear bandits with general nonlinear constraints. The objective is to maximize the expected cumulative reward over horizon $T$ subject to a set of constraints in each round $\tau\leq T$. We propose a pessimistic-optimistic algorithm for this problem, which is efficient in two aspects. First, the algorithm yields $\tilde{\cal O}\left(\left(\frac{K^{0.75}}{\delta}+d\right)\sqrt{\tau}\right)$ (pseudo) regret in round $\tau\leq T,$ where $K$ is the number of constraints, $d$ is the dimension of the reward feature space, and $\delta$ is a Slater's constant; and {\em zero} constraint violation in any round $\tau>\tau',$ where $\tau'$ is {\em independent} of horizon $T.$ Second, the algorithm is computationally efficient. Our algorithm is based on the primal-dual approach in optimization and includes two components. The primal component is similar to unconstrained stochastic linear bandits (our algorithm uses the linear upper confidence bound algorithm (LinUCB)). The computational complexity of the dual component depends on the number of constraints, but is independent of the sizes of the contextual space, the action space, and the feature space. Thus, the computational complexity of our algorithm is similar to LinUCB for unconstrained stochastic linear bandits.
| null |
Efficiently Learning One Hidden Layer ReLU Networks From Queries
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca4b5656b7e193e6bb9064c672ac8dce-Abstract.html
|
Sitan Chen, Adam Klivans, Raghu Meka
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca4b5656b7e193e6bb9064c672ac8dce-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13467-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca4b5656b7e193e6bb9064c672ac8dce-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=oZg-aOyHL-h
|
https://papers.nips.cc/paper_files/paper/2021/file/ca4b5656b7e193e6bb9064c672ac8dce-Supplemental.pdf
|
While the problem of PAC learning neural networks from samples has received considerable attention in recent years, in certain settings like model extraction attacks, it is reasonable to imagine having more than just the ability to observe random labeled examples. Motivated by this, we consider the following problem: given \emph{black-box query access} to a neural network $F$, recover $F$ up to some error. Formally, we show that if $F$ is an arbitrary one hidden layer neural network with ReLU activations, there is an algorithm with query complexity and runtime polynomial in all parameters which outputs a network $F’$ achieving low square loss relative to $F$ with respect to the Gaussian measure. While a number of works in the security literature have proposed and empirically demonstrated the effectiveness of certain algorithms for this problem, ours is to the best of our knowledge the first provable guarantee in this vein.
| null |
Learning Nonparametric Volterra Kernels with Gaussian Processes
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca5fbbbddd0c0ff6c01f782c60c9d1b5-Abstract.html
|
Magnus Ross, Michael T Smith, Mauricio Álvarez
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca5fbbbddd0c0ff6c01f782c60c9d1b5-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13468-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca5fbbbddd0c0ff6c01f782c60c9d1b5-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=FyI2-YoHHd
|
https://papers.nips.cc/paper_files/paper/2021/file/ca5fbbbddd0c0ff6c01f782c60c9d1b5-Supplemental.pdf
|
This paper introduces a method for the nonparametric Bayesian learning of nonlinear operators, through the use of the Volterra series with kernels represented using Gaussian processes (GPs), which we term the nonparametric Volterra kernels model (NVKM). When the input function to the operator is unobserved and has a GP prior, the NVKM constitutes a powerful method for both single and multiple output regression, and can be viewed as a nonlinear and nonparametric latent force model. When the input function is observed, the NVKM can be used to perform Bayesian system identification. We use recent advances in efficient sampling of explicit functions from GPs to map process realisations through the Volterra series without resorting to numerical integration, allowing scalability through doubly stochastic variational inference, and avoiding the need for Gaussian approximations of the output processes. We demonstrate the performance of the model for both multiple output regression and system identification using standard benchmarks.
| null |
DiBS: Differentiable Bayesian Structure Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca6ab34959489659f8c3776aaf1f8efd-Abstract.html
|
Lars Lorch, Jonas Rothfuss, Bernhard Schölkopf, Andreas Krause
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca6ab34959489659f8c3776aaf1f8efd-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13469-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca6ab34959489659f8c3776aaf1f8efd-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=YqYt54gU-XV
|
https://papers.nips.cc/paper_files/paper/2021/file/ca6ab34959489659f8c3776aaf1f8efd-Supplemental.pdf
|
Bayesian structure learning allows inferring Bayesian network structure from data while reasoning about the epistemic uncertainty---a key element towards enabling active causal discovery and designing interventions in real world systems. In this work, we propose a general, fully differentiable framework for Bayesian structure learning (DiBS) that operates in the continuous space of a latent probabilistic graph representation. Contrary to existing work, DiBS is agnostic to the form of the local conditional distributions and allows for joint posterior inference of both the graph structure and the conditional distribution parameters. This makes our formulation directly applicable to posterior inference of nonstandard Bayesian network models, e.g., with nonlinear dependencies encoded by neural networks. Using DiBS, we devise an efficient, general purpose variational inference method for approximating distributions over structural models. In evaluations on simulated and real-world data, our method significantly outperforms related approaches to joint posterior inference.
| null |
Nonparametric estimation of continuous DPPs with kernel methods
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca8a2d76a5bcc212226417361a5f0740-Abstract.html
|
Michaël Fanuel, Rémi Bardenet
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca8a2d76a5bcc212226417361a5f0740-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13470-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca8a2d76a5bcc212226417361a5f0740-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=MGHO3xLMohC
|
https://papers.nips.cc/paper_files/paper/2021/file/ca8a2d76a5bcc212226417361a5f0740-Supplemental.pdf
|
Determinantal Point Process (DPPs) are statistical models for repulsive point patterns. Both sampling and inference are tractable for DPPs, a rare feature among models with negative dependence that explains their popularity in machine learning and spatial statistics. Parametric and nonparametric inference methods have been proposed in the finite case, i.e. when the point patterns live in a finite ground set. In the continuous case, only parametric methods have been investigated, while nonparametric maximum likelihood for DPPs -- an optimization problem over trace-class operators -- has remained an open question. In this paper, we show that a restricted version of this maximum likelihood (MLE) problem falls within the scope of a recent representer theorem for nonnegative functions in an RKHS. This leads to a finite-dimensional problem, with strong statistical ties to the original MLE. Moreover, we propose, analyze, and demonstrate a fixed point algorithm to solve this finite-dimensional problem. Finally, we also provide a controlled estimate of the correlation kernel of the DPP, thus providing more interpretability.
| null |
FINE Samples for Learning with Noisy Labels
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca91c5464e73d3066825362c3093a45f-Abstract.html
|
Taehyeon Kim, Jongwoo Ko, sangwook Cho, JinHwan Choi, Se-Young Yun
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca91c5464e73d3066825362c3093a45f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13471-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca91c5464e73d3066825362c3093a45f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=QZpx42n0BWr
|
https://papers.nips.cc/paper_files/paper/2021/file/ca91c5464e73d3066825362c3093a45f-Supplemental.pdf
|
Modern deep neural networks (DNNs) become frail when the datasets contain noisy (incorrect) class labels. Robust techniques in the presence of noisy labels can be categorized into two folds: developing noise-robust functions or using noise-cleansing methods by detecting the noisy data. Recently, noise-cleansing methods have been considered as the most competitive noisy-label learning algorithms. Despite their success, their noisy label detectors are often based on heuristics more than a theory, requiring a robust classifier to predict the noisy data with loss values. In this paper, we propose a novel detector for filtering label noise. Unlike most existing methods, we focus on each data's latent representation dynamics and measure the alignment between the latent distribution and each representation using the eigen decomposition of the data gram matrix. Our framework, coined as filtering noisy instances via their eigenvectors (FINE), provides a robust detector with derivative-free simple methods having theoretical guarantees. Under our framework, we propose three applications of the FINE: sample-selection approach, semi-supervised learning approach, and collaboration with noise-robust loss functions. Experimental results show that the proposed methods consistently outperform corresponding baselines for all three applications on various benchmark datasets.
| null |
Residual2Vec: Debiasing graph embedding with random graphs
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca9541826e97c4530b07dda2eba0e013-Abstract.html
|
Sadamori Kojaku, Jisung Yoon, Isabel Constantino, Yong-Yeol Ahn
|
https://papers.nips.cc/paper_files/paper/2021/hash/ca9541826e97c4530b07dda2eba0e013-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13472-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ca9541826e97c4530b07dda2eba0e013-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=AHFfYwM7WGP
| null |
Graph embedding maps a graph into a convenient vector-space representation for graph analysis and machine learning applications. Many graph embedding methods hinge on a sampling of context nodes based on random walks. However, random walks can be a biased sampler due to the structural properties of graphs. Most notably, random walks are biased by the degree of each node, where a node is sampled proportionally to its degree. The implication of such biases has not been clear, particularly in the context of graph representation learning. Here, we investigate the impact of the random walks' bias on graph embedding and propose residual2vec, a general graph embedding method that can debias various structural biases in graphs by using random graphs. We demonstrate that this debiasing not only improves link prediction and clustering performance but also allows us to explicitly model salient structural properties in graph embedding.
| null |
Benign Overfitting in Multiclass Classification: All Roads Lead to Interpolation
|
https://papers.nips.cc/paper_files/paper/2021/hash/caaa29eab72b231b0af62fbdff89bfce-Abstract.html
|
Ke Wang, Vidya Muthukumar, Christos Thrampoulidis
|
https://papers.nips.cc/paper_files/paper/2021/hash/caaa29eab72b231b0af62fbdff89bfce-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13473-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/caaa29eab72b231b0af62fbdff89bfce-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=0O5jpovbdHO
|
https://papers.nips.cc/paper_files/paper/2021/file/caaa29eab72b231b0af62fbdff89bfce-Supplemental.pdf
|
The growing literature on "benign overfitting" in overparameterized models has been mostly restricted to regression or binary classification settings; however, most success stories of modern machine learning have been recorded in multiclass settings. Motivated by this discrepancy, we study benign overfitting in multiclass linear classification. Specifically, we consider the following popular training algorithms on separable data: (i) empirical risk minimization (ERM) with cross-entropy loss, which converges to the multiclass support vector machine (SVM) solution; (ii) ERM with least-squares loss, which converges to the min-norm interpolating (MNI) solution; and, (iii) the one-vs-all SVM classifier. Our first key finding is that under a simple sufficient condition, all three algorithms lead to classifiers that interpolate the training data and have equal accuracy. When the data is generated from Gaussian mixtures or a multinomial logistic model, this condition holds under high enough effective overparameterization. Second, we derive novel error bounds on the accuracy of the MNI classifier, thereby showing that all three training algorithms lead to benign overfitting under sufficient overparameterization. Ultimately, our analysis shows that good generalization is possible for SVM solutions beyond the realm in which typical margin-based bounds apply.
| null |
Instance-Dependent Bounds for Zeroth-order Lipschitz Optimization with Error Certificates
|
https://papers.nips.cc/paper_files/paper/2021/hash/cacbf64b8a464fa1974da1eb0aa92851-Abstract.html
|
Francois Bachoc, Tom Cesari, Sébastien Gerchinovitz
|
https://papers.nips.cc/paper_files/paper/2021/hash/cacbf64b8a464fa1974da1eb0aa92851-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13474-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cacbf64b8a464fa1974da1eb0aa92851-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=yrqn9rQO2YT
|
https://papers.nips.cc/paper_files/paper/2021/file/cacbf64b8a464fa1974da1eb0aa92851-Supplemental.pdf
|
We study the problem of zeroth-order (black-box) optimization of a Lipschitz function $f$ defined on a compact subset $\mathcal{X}$ of $\mathbb{R}^d$, with the additional constraint that algorithms must certify the accuracy of their recommendations. We characterize the optimal number of evaluations of any Lipschitz function $f$ to find and certify an approximate maximizer of $f$ at accuracy $\varepsilon$. Under a weak assumption on $\mathcal{X}$, this optimal sample complexity is shown to be nearly proportional to the integral $\int_{\mathcal{X}} \mathrm{d}\boldsymbol{x}/( \max(f) - f(\boldsymbol{x}) + \varepsilon )^d$. This result, which was only (and partially) known in dimension $d=1$, solves an open problem dating back to 1991. In terms of techniques, our upper bound relies on a packing bound by Bouttier et al. (2020) for the Piyavskii-Shubert algorithm that we link to the above integral. We also show that a certified version of the computationally tractable DOO algorithm matches these packing and integral bounds. Our instance-dependent lower bound differs from traditional worst-case lower bounds in the Lipschitz setting and relies on a local worst-case analysis that could likely prove useful for other learning tasks.
| null |
Training Neural Networks with Fixed Sparse Masks
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb2653f548f8709598e8b5156738cc51-Abstract.html
|
Yi-Lin Sung, Varun Nair, Colin A. Raffel
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb2653f548f8709598e8b5156738cc51-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13475-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cb2653f548f8709598e8b5156738cc51-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=Uwh-v1HSw-x
|
https://papers.nips.cc/paper_files/paper/2021/file/cb2653f548f8709598e8b5156738cc51-Supplemental.zip
|
During typical gradient-based training of deep neural networks, all of the model's parameters are updated at each iteration. Recent work has shown that it is possible to update only a small subset of the model's parameters during training, which can alleviate storage and communication requirements. In this paper, we show that it is possible to induce a fixed sparse mask on the model’s parameters that selects a subset to update over many iterations. Our method constructs the mask out of the $k$ parameters with the largest Fisher information as a simple approximation as to which parameters are most important for the task at hand. In experiments on parameter-efficient transfer learning and distributed training, we show that our approach matches or exceeds the performance of other methods for training with sparse updates while being more efficient in terms of memory usage and communication costs. We release our code publicly to promote further applications of our approach.
| null |
VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb3213ada48302953cb0f166464ab356-Abstract.html
|
Hassan Akbari, Liangzhe Yuan, Rui Qian, Wei-Hong Chuang, Shih-Fu Chang, Yin Cui, Boqing Gong
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb3213ada48302953cb0f166464ab356-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13476-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cb3213ada48302953cb0f166464ab356-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=RzYrn625bu8
|
https://papers.nips.cc/paper_files/paper/2021/file/cb3213ada48302953cb0f166464ab356-Supplemental.pdf
|
We present a framework for learning multimodal representations from unlabeled data using convolution-free Transformer architectures. Specifically, our Video-Audio-Text Transformer (VATT) takes raw signals as inputs and extracts multimodal representations that are rich enough to benefit a variety of downstream tasks. We train VATT end-to-end from scratch using multimodal contrastive losses and evaluate its performance by the downstream tasks of video action recognition, audio event classification, image classification, and text-to-video retrieval. Furthermore, we study a modality-agnostic single-backbone Transformer by sharing weights among the three modalities. We show that the convolution-free VATT outperforms state-of-the-art ConvNet-based architectures in the downstream tasks. Especially, VATT's vision Transformer achieves the top-1 accuracy of 82.1% on Kinetics-400, 83.6% on Kinetics-600, 72.7% on Kinetics-700, and 41.1% on Moments in Time, new records while avoiding supervised pre-training. Transferring to image classification leads to 78.7% top-1 accuracy on ImageNet compared to 64.7% by training the same Transformer from scratch, showing the generalizability of our model despite the domain gap between videos and images. VATT's audio Transformer also sets a new record on waveform-based audio event recognition by achieving the mAP of 39.4% on AudioSet without any supervised pre-training.
| null |
Analyzing the Generalization Capability of SGLD Using Properties of Gaussian Channels
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb77649f5d53798edfa0ff40dae46322-Abstract.html
|
Hao Wang, Yizhe Huang, Rui Gao, Flavio Calmon
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb77649f5d53798edfa0ff40dae46322-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13477-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cb77649f5d53798edfa0ff40dae46322-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=t-7Jx48oaG
|
https://papers.nips.cc/paper_files/paper/2021/file/cb77649f5d53798edfa0ff40dae46322-Supplemental.pdf
|
Optimization is a key component for training machine learning models and has a strong impact on their generalization. In this paper, we consider a particular optimization method---the stochastic gradient Langevin dynamics (SGLD) algorithm---and investigate the generalization of models trained by SGLD. We derive a new generalization bound by connecting SGLD with Gaussian channels found in information and communication theory. Our bound can be computed from the training data and incorporates the variance of gradients for quantifying a particular kind of "sharpness" of the loss landscape. We also consider a closely related algorithm with SGLD, namely differentially private SGD (DP-SGD). We prove that the generalization capability of DP-SGD can be amplified by iteration. Specifically, our bound can be sharpened by including a time-decaying factor if the DP-SGD algorithm outputs the last iterate while keeping other iterates hidden. This decay factor enables the contribution of early iterations to our bound to reduce with time and is established by strong data processing inequalities---a fundamental tool in information theory. We demonstrate our bound through numerical experiments, showing that it can predict the behavior of the true generalization gap.
| null |
Learning to Schedule Heuristics in Branch and Bound
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb7c403aa312160380010ee3dd4bfc53-Abstract.html
|
Antonia Chmiela, Elias Khalil, Ambros Gleixner, Andrea Lodi, Sebastian Pokutta
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb7c403aa312160380010ee3dd4bfc53-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13478-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cb7c403aa312160380010ee3dd4bfc53-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=fEImgFxKU63
|
https://papers.nips.cc/paper_files/paper/2021/file/cb7c403aa312160380010ee3dd4bfc53-Supplemental.pdf
|
Primal heuristics play a crucial role in exact solvers for Mixed Integer Programming (MIP). While solvers are guaranteed to find optimal solutions given sufficient time, real-world applications typically require finding good solutions early on in the search to enable fast decision-making. While much of MIP research focuses on designing effective heuristics, the question of how to manage multiple MIP heuristics in a solver has not received equal attention. Generally, solvers follow hard-coded rules derived from empirical testing on broad sets of instances. Since the performance of heuristics is problem-dependent, using these general rules for a particular problem might not yield the best performance. In this work, we propose the first data-driven framework for scheduling heuristics in an exact MIP solver. By learning from data describing the performance of primal heuristics, we obtain a problem-specific schedule of heuristics that collectively find many solutions at minimal cost. We formalize the learning task and propose an efficient algorithm for computing such a schedule. Compared to the default settings of a state-of-the-art academic MIP solver, we are able to reduce the average primal integral by up to 49% on two classes of challenging instances.
| null |
On Training Implicit Models
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb8da6767461f2812ae4290eac7cbc42-Abstract.html
|
Zhengyang Geng, Xin-Yu Zhang, Shaojie Bai, Yisen Wang, Zhouchen Lin
|
https://papers.nips.cc/paper_files/paper/2021/hash/cb8da6767461f2812ae4290eac7cbc42-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13479-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cb8da6767461f2812ae4290eac7cbc42-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=3EwcMzmUbNd
|
https://papers.nips.cc/paper_files/paper/2021/file/cb8da6767461f2812ae4290eac7cbc42-Supplemental.zip
|
This paper focuses on training implicit models of infinite layers. Specifically, previous works employ implicit differentiation and solve the exact gradient for the backward propagation. However, is it necessary to compute such an exact but expensive gradient for training? In this work, we propose a novel gradient estimate for implicit models, named phantom gradient, that 1) forgoes the costly computation of the exact gradient; and 2) provides an update direction empirically preferable to the implicit model training. We theoretically analyze the condition under which an ascent direction of the loss landscape could be found and provide two specific instantiations of the phantom gradient based on the damped unrolling and Neumann series. Experiments on large-scale tasks demonstrate that these lightweight phantom gradients significantly accelerate the backward passes in training implicit models by roughly 1.7 $\times$ and even boost the performance over approaches based on the exact gradient on ImageNet.
| null |
MLP-Mixer: An all-MLP Architecture for Vision
|
https://papers.nips.cc/paper_files/paper/2021/hash/cba0a4ee5ccd02fda0fe3f9a3e7b89fe-Abstract.html
|
Ilya O. Tolstikhin, Neil Houlsby, Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Thomas Unterthiner, Jessica Yung, Andreas Steiner, Daniel Keysers, Jakob Uszkoreit, Mario Lucic, Alexey Dosovitskiy
|
https://papers.nips.cc/paper_files/paper/2021/hash/cba0a4ee5ccd02fda0fe3f9a3e7b89fe-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13480-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cba0a4ee5ccd02fda0fe3f9a3e7b89fe-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=EI2KOXKdnP
|
https://papers.nips.cc/paper_files/paper/2021/file/cba0a4ee5ccd02fda0fe3f9a3e7b89fe-Supplemental.pdf
|
Convolutional Neural Networks (CNNs) are the go-to model for computer vision. Recently, attention-based networks, such as the Vision Transformer, have also become popular. In this paper we show that while convolutions and attention are both sufficient for good performance, neither of them are necessary. We present MLP-Mixer, an architecture based exclusively on multi-layer perceptrons (MLPs). MLP-Mixer contains two types of layers: one with MLPs applied independently to image patches (i.e. "mixing" the per-location features), and one with MLPs applied across patches (i.e. "mixing" spatial information). When trained on large datasets, or with modern regularization schemes, MLP-Mixer attains competitive scores on image classification benchmarks, with pre-training and inference cost comparable to state-of-the-art models. We hope that these results spark further research beyond the realms of well established CNNs and Transformers.
| null |
A Framework to Learn with Interpretation
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbb6a3b884f4f88b3a8e3d44c636cbd8-Abstract.html
|
Jayneel Parekh, Pavlo Mozharovskyi, Florence d'Alché-Buc
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbb6a3b884f4f88b3a8e3d44c636cbd8-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13481-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cbb6a3b884f4f88b3a8e3d44c636cbd8-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=k_w-RCJ9kMw
|
https://papers.nips.cc/paper_files/paper/2021/file/cbb6a3b884f4f88b3a8e3d44c636cbd8-Supplemental.pdf
|
To tackle interpretability in deep learning, we present a novel framework to jointly learn a predictive model and its associated interpretation model. The interpreter provides both local and global interpretability about the predictive model in terms of human-understandable high level attribute functions, with minimal loss of accuracy. This is achieved by a dedicated architecture and well chosen regularization penalties. We seek for a small-size dictionary of high level attribute functions that take as inputs the outputs of selected hidden layers and whose outputs feed a linear classifier. We impose strong conciseness on the activation of attributes with an entropy-based criterion while enforcing fidelity to both inputs and outputs of the predictive model. A detailed pipeline to visualize the learnt features is also developed. Moreover, besides generating interpretable models by design, our approach can be specialized to provide post-hoc interpretations for a pre-trained neural network. We validate our approach against several state-of-the-art methods on multiple datasets and show its efficacy on both kinds of tasks.
| null |
One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbcb58ac2e496207586df2854b17995f-Abstract.html
|
Jiun Tian Hoe, Kam Woh Ng, Tianyu Zhang, Chee Seng Chan, Yi-Zhe Song, Tao Xiang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbcb58ac2e496207586df2854b17995f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13482-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cbcb58ac2e496207586df2854b17995f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=2pJZSVcSZz
|
https://papers.nips.cc/paper_files/paper/2021/file/cbcb58ac2e496207586df2854b17995f-Supplemental.pdf
|
A deep hashing model typically has two main learning objectives: to make the learned binary hash codes discriminative and to minimize a quantization error. With further constraints such as bit balance and code orthogonality, it is not uncommon for existing models to employ a large number (>4) of losses. This leads to difficulties in model training and subsequently impedes their effectiveness. In this work, we propose a novel deep hashing model with only $\textit{a single learning objective}$. Specifically, we show that maximizing the cosine similarity between the continuous codes and their corresponding $\textit{binary orthogonal codes}$ can ensure both hash code discriminativeness and quantization error minimization. Further, with this learning objective, code balancing can be achieved by simply using a Batch Normalization (BN) layer and multi-label classification is also straightforward with label smoothing. The result is a one-loss deep hashing model that removes all the hassles of tuning the weights of various losses. Importantly, extensive experiments show that our model is highly effective, outperforming the state-of-the-art multi-loss hashing models on three large-scale instance retrieval benchmarks, often by significant margins.
| null |
Fast and accurate randomized algorithms for low-rank tensor decompositions
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbef46321026d8404bc3216d4774c8a9-Abstract.html
|
Linjian Ma, Edgar Solomonik
|
https://papers.nips.cc/paper_files/paper/2021/hash/cbef46321026d8404bc3216d4774c8a9-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13483-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cbef46321026d8404bc3216d4774c8a9-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=B4szfz7W7LU
|
https://papers.nips.cc/paper_files/paper/2021/file/cbef46321026d8404bc3216d4774c8a9-Supplemental.pdf
|
Low-rank Tucker and CP tensor decompositions are powerful tools in data analytics. The widely used alternating least squares (ALS) method, which solves a sequence of over-determined least squares subproblems, is costly for large and sparse tensors. We propose a fast and accurate sketched ALS algorithm for Tucker decomposition, which solves a sequence of sketched rank-constrained linear least squares subproblems. Theoretical sketch size upper bounds are provided to achieve $O(\epsilon)$ relative error for each subproblem with two sketching techniques, TensorSketch and leverage score sampling. Experimental results show that this new ALS algorithm, combined with a new initialization scheme based on the randomized range finder, yields decomposition accuracy comparable to the standard higher-order orthogonal iteration (HOOI) algorithm. The new algorithm achieves up to $22.0\%$ relative decomposition residual improvement compared to the state-of-the-art sketched randomized algorithm for Tucker decomposition of various synthetic and real datasets. This Tucker-ALS algorithm is further used to accelerate CP decomposition, by using randomized Tucker compression followed by CP decomposition of the Tucker core tensor. Experimental results show that this algorithm not only converges faster, but also yields more accurate CP decompositions.
| null |
Communication-efficient SGD: From Local SGD to One-Shot Averaging
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc06a6150b92e17dd3076a0f0f9d2af4-Abstract.html
|
Artin Spiridonoff, Alex Olshevsky, Yannis Paschalidis
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc06a6150b92e17dd3076a0f0f9d2af4-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13484-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc06a6150b92e17dd3076a0f0f9d2af4-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=UpfqzQtZ58
|
https://papers.nips.cc/paper_files/paper/2021/file/cc06a6150b92e17dd3076a0f0f9d2af4-Supplemental.zip
|
We consider speeding up stochastic gradient descent (SGD) by parallelizing it across multiple workers. We assume the same data set is shared among $N$ workers, who can take SGD steps and coordinate with a central server. While it is possible to obtain a linear reduction in the variance by averaging all the stochastic gradients at every step, this requires a lot of communication between the workers and the server, which can dramatically reduce the gains from parallelism.The Local SGD method, proposed and analyzed in the earlier literature, suggests machines should make many local steps between such communications. While the initial analysis of Local SGD showed it needs $\Omega ( \sqrt{T} )$ communications for $T$ local gradient steps in order for the error to scale proportionately to $1/(NT)$, this has been successively improved in a string of papers, with the state of the art requiring $\Omega \left( N \left( \mbox{ poly} (\log T) \right) \right)$ communications. In this paper, we suggest a Local SGD scheme that communicates less overall by communicating less frequently as the number of iterations grows. Our analysis shows that this can achieve an error that scales as $1/(NT)$ with a number of communications that is completely independent of $T$. In particular, we show that $\Omega(N)$ communications are sufficient. Empirical evidence suggests this bound is close to tight as we further show that $\sqrt{N}$ or $N^{3/4}$ communications fail to achieve linear speed-up in simulations. Moreover, we show that under mild assumptions, the main of which is twice differentiability on any neighborhood of the optimal solution, one-shot averaging which only uses a single round of communication can also achieve the optimal convergence rate asymptotically.
| null |
Memory Efficient Meta-Learning with Large Images
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc1aa436277138f61cda703991069eaf-Abstract.html
|
John Bronskill, Daniela Massiceti, Massimiliano Patacchiola, Katja Hofmann, Sebastian Nowozin, Richard Turner
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc1aa436277138f61cda703991069eaf-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13485-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc1aa436277138f61cda703991069eaf-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=x2pF7Tt_S5u
|
https://papers.nips.cc/paper_files/paper/2021/file/cc1aa436277138f61cda703991069eaf-Supplemental.pdf
|
Meta learning approaches to few-shot classification are computationally efficient at test time, requiring just a few optimization steps or single forward pass to learn a new task, but they remain highly memory-intensive to train. This limitation arises because a task's entire support set, which can contain up to 1000 images, must be processed before an optimization step can be taken. Harnessing the performance gains offered by large images thus requires either parallelizing the meta-learner across multiple GPUs, which may not be available, or trade-offs between task and image size when memory constraints apply. We improve on both options by proposing LITE, a general and memory efficient episodic training scheme that enables meta-training on large tasks composed of large images on a single GPU. We achieve this by observing that the gradients for a task can be decomposed into a sum of gradients over the task's training images. This enables us to perform a forward pass on a task's entire training set but realize significant memory savings by back-propagating only a random subset of these images which we show is an unbiased approximation of the full gradient. We use LITE to train meta-learners and demonstrate new state-of-the-art accuracy on the real-world ORBIT benchmark and 3 of the 4 parts of the challenging VTAB+MD benchmark relative to leading meta-learners. LITE also enables meta-learners to be competitive with transfer learning approaches but at a fraction of the test-time computational cost, thus serving as a counterpoint to the recent narrative that transfer learning is all you need for few-shot classification.
| null |
On the Power of Differentiable Learning versus PAC and SQ Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc225865b743ecc91c4743259813f604-Abstract.html
|
Emmanuel Abbe, Pritish Kamath, Eran Malach, Colin Sandon, Nathan Srebro
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc225865b743ecc91c4743259813f604-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13486-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc225865b743ecc91c4743259813f604-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=WYrC0Aentah
|
https://papers.nips.cc/paper_files/paper/2021/file/cc225865b743ecc91c4743259813f604-Supplemental.pdf
|
We study the power of learning via mini-batch stochastic gradient descent (SGD) on the loss of a differentiable model or neural network, and ask what learning problems can be learnt using this paradigm. We show that SGD can always simulate learning with statistical queries (SQ), but its ability to go beyond that depends on the precision $\rho$ of the gradients and the minibatch size $b$. With fine enough precision relative to minibatch size, namely when $b \rho$ is small enough, SGD can go beyond SQ learning and simulate any sample-based learning algorithm and thus its learning power is equivalent to that of PAC learning; this extends prior work that achieved this result for $b=1$. Moreover, with polynomially many bits of precision (i.e. when $\rho$ is exponentially small), SGD can simulate PAC learning regardless of the batch size. On the other hand, when $b \rho^2$ is large enough, the power of SGD is equivalent to that of SQ learning.
| null |
Can we globally optimize cross-validation loss? Quasiconvexity in ridge regression
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc298d5bc587e1b650f80e10449ee9d5-Abstract.html
|
Will Stephenson, Zachary Frangella, Madeleine Udell, Tamara Broderick
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc298d5bc587e1b650f80e10449ee9d5-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13487-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc298d5bc587e1b650f80e10449ee9d5-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=4Il6i0jdrvP
|
https://papers.nips.cc/paper_files/paper/2021/file/cc298d5bc587e1b650f80e10449ee9d5-Supplemental.pdf
|
Models like LASSO and ridge regression are extensively used in practice due to their interpretability, ease of use, and strong theoretical guarantees. Cross-validation (CV) is widely used for hyperparameter tuning in these models, but do practical methods minimize the true out-of-sample loss? A recent line of research promises to show that the optimum of the CV loss matches the optimum of the out-of-sample loss (possibly after simple corrections). It remains to show how tractable it is to minimize the CV loss.In the present paper, we show that, in the case of ridge regression, the CV loss may fail to be quasiconvex and thus may have multiple local optima. We can guarantee that the CV loss is quasiconvex in at least one case: when the spectrum of the covariate matrix is nearly flat and the noise in the observed responses is not too high. More generally, we show that quasiconvexity status is independent of many properties of the observed data (response norm, covariate-matrix right singular vectors and singular-value scaling) and has a complex dependence on the few that remain. We empirically confirm our theory using simulated experiments.
| null |
Adaptive Proximal Gradient Methods for Structured Neural Networks
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc3f5463bc4d26bc38eadc8bcffbc654-Abstract.html
|
Jihun Yun, Aurelie C. Lozano, Eunho Yang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc3f5463bc4d26bc38eadc8bcffbc654-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13488-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc3f5463bc4d26bc38eadc8bcffbc654-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=Qijzj3WqUl3
|
https://papers.nips.cc/paper_files/paper/2021/file/cc3f5463bc4d26bc38eadc8bcffbc654-Supplemental.pdf
|
We consider the training of structured neural networks where the regularizer can be non-smooth and possibly non-convex. While popular machine learning libraries have resorted to stochastic (adaptive) subgradient approaches, the use of proximal gradient methods in the stochastic setting has been little explored and warrants further study, in particular regarding the incorporation of adaptivity. Towards this goal, we present a general framework of stochastic proximal gradient descent methods that allows for arbitrary positive preconditioners and lower semi-continuous regularizers. We derive two important instances of our framework: (i) the first proximal version of \textsc{Adam}, one of the most popular adaptive SGD algorithm, and (ii) a revised version of ProxQuant for quantization-specific regularizers, which improves upon the original approach by incorporating the effect of preconditioners in the proximal mapping computations. We provide convergence guarantees for our framework and show that adaptive gradient methods can have faster convergence in terms of constant than vanilla SGD for sparse data. Lastly, we demonstrate the superiority of stochastic proximal methods compared to subgradient-based approaches via extensive experiments. Interestingly, our results indicate that the benefit of proximal approaches over sub-gradient counterparts is more pronounced for non-convex regularizers than for convex ones.
| null |
Discovering and Achieving Goals via World Models
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc4af25fa9d2d5c953496579b75f6f6c-Abstract.html
|
Russell Mendonca, Oleh Rybkin, Kostas Daniilidis, Danijar Hafner, Deepak Pathak
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc4af25fa9d2d5c953496579b75f6f6c-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13489-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc4af25fa9d2d5c953496579b75f6f6c-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=6vWuYzkp8d
|
https://papers.nips.cc/paper_files/paper/2021/file/cc4af25fa9d2d5c953496579b75f6f6c-Supplemental.pdf
|
How can artificial agents learn to solve many diverse tasks in complex visual environments without any supervision? We decompose this question into two challenges: discovering new goals and learning to reliably achieve them. Our proposed agent, Latent Explorer Achiever (LEXA), addresses both challenges by learning a world model from image inputs and using it to train an explorer and an achiever policy via imagined rollouts. Unlike prior methods that explore by reaching previously visited states, the explorer plans to discover unseen surprising states through foresight, which are then used as diverse targets for the achiever to practice. After the unsupervised phase, LEXA solves tasks specified as goal images zero-shot without any additional learning. LEXA substantially outperforms previous approaches to unsupervised goal reaching, both on prior benchmarks and on a new challenging benchmark with 40 test tasks spanning across four robotic manipulation and locomotion domains. LEXA further achieves goals that require interacting with multiple objects in sequence. Project page: https://orybkin.github.io/lexa/
| null |
Understanding and Improving Early Stopping for Learning with Noisy Labels
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc7e2b878868cbae992d1fb743995d8f-Abstract.html
|
Yingbin Bai, Erkun Yang, Bo Han, Yanhua Yang, Jiatong Li, Yinian Mao, Gang Niu, Tongliang Liu
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc7e2b878868cbae992d1fb743995d8f-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13490-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc7e2b878868cbae992d1fb743995d8f-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=KbV-UZRKb3g
|
https://papers.nips.cc/paper_files/paper/2021/file/cc7e2b878868cbae992d1fb743995d8f-Supplemental.pdf
|
The memorization effect of deep neural network (DNN) plays a pivotal role in many state-of-the-art label-noise learning methods. To exploit this property, the early stopping trick, which stops the optimization at the early stage of training, is usually adopted. Current methods generally decide the early stopping point by considering a DNN as a whole. However, a DNN can be considered as a composition of a series of layers, and we find that the latter layers in a DNN are much more sensitive to label noise, while their former counterparts are quite robust. Therefore, selecting a stopping point for the whole network may make different DNN layers antagonistically affect each other, thus degrading the final performance. In this paper, we propose to separate a DNN into different parts and progressively train them to address this problem. Instead of the early stopping which trains a whole DNN all at once, we initially train former DNN layers by optimizing the DNN with a relatively large number of epochs. During training, we progressively train the latter DNN layers by using a smaller number of epochs with the preceding layers fixed to counteract the impact of noisy labels. We term the proposed method as progressive early stopping (PES). Despite its simplicity, compared with the traditional early stopping, PES can help to obtain more promising and stable results. Furthermore, by combining PES with existing approaches on noisy label training, we achieve state-of-the-art performance on image classification benchmarks. The code is made public at https://github.com/tmllab/PES.
| null |
Distributionally Robust Imitation Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc8090c4d2791cdd9cd2cb3c24296190-Abstract.html
|
Mohammad Ali Bashiri, Brian Ziebart, Xinhua Zhang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc8090c4d2791cdd9cd2cb3c24296190-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13491-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc8090c4d2791cdd9cd2cb3c24296190-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=PJEPtZmw-SQ
|
https://papers.nips.cc/paper_files/paper/2021/file/cc8090c4d2791cdd9cd2cb3c24296190-Supplemental.pdf
|
We consider the imitation learning problem of learning a policy in a Markov Decision Process (MDP) setting where the reward function is not given, but demonstrations from experts are available. Although the goal of imitation learning is to learn a policy that produces behaviors nearly as good as the experts’ for a desired task, assumptions of consistent optimality for demonstrated behaviors are often violated in practice. Finding a policy that is distributionally robust against noisy demonstrations based on an adversarial construction potentially solves this problem by avoiding optimistic generalizations of the demonstrated data. This paper studies Distributionally Robust Imitation Learning (DRoIL) and establishes a close connection between DRoIL and Maximum Entropy Inverse Reinforcement Learning. We show that DRoIL can be seen as a framework that maximizes a generalized concept of entropy. We develop a novel approach to transform the objective function into a convex optimization problem over a polynomial number of variables for a class of loss functions that are additive over state and action spaces. Our approach lets us optimize both stationary and non-stationary policies and, unlike prevalent previous methods, it does not require repeatedly solving an inner reinforcement learning problem. We experimentally show the significant benefits of DRoIL’s new optimization method on synthetic data and a highway driving environment.
| null |
On the Power of Edge Independent Graph Models
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc9b3c69b56df284846bf2432f1cba90-Abstract.html
|
Sudhanshu Chanpuriya, Cameron Musco, Konstantinos Sotiropoulos, Charalampos Tsourakakis
|
https://papers.nips.cc/paper_files/paper/2021/hash/cc9b3c69b56df284846bf2432f1cba90-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13492-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cc9b3c69b56df284846bf2432f1cba90-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=OrPraBRj45z
|
https://papers.nips.cc/paper_files/paper/2021/file/cc9b3c69b56df284846bf2432f1cba90-Supplemental.pdf
|
Why do many modern neural-network-based graph generative models fail to reproduce typical real-world network characteristics, such as high triangle density? In this work we study the limitations of $edge\ independent\ random\ graph\ models$, in which each edge is added to the graph independently with some probability. Such models include both the classic Erdos-Renyi and stochastic block models, as well as modern generative models such as NetGAN, variational graph autoencoders, and CELL. We prove that subject to a $bounded\ overlap$ condition, which ensures that the model does not simply memorize a single graph, edge independent models are inherently limited in their ability to generate graphs with high triangle and other subgraph densities. Notably, such high densities are known to appear in real-world social networks and other graphs. We complement our negative results with a simple generative model that balances overlap and accuracy, performing comparably to more complex models in reconstructing many graph statistics.
| null |
Stochastic Online Linear Regression: the Forward Algorithm to Replace Ridge
|
https://papers.nips.cc/paper_files/paper/2021/hash/cca289d2a4acd14c1cd9a84ffb41dd29-Abstract.html
|
Reda Ouhamma, Odalric-Ambrym Maillard, Vianney Perchet
|
https://papers.nips.cc/paper_files/paper/2021/hash/cca289d2a4acd14c1cd9a84ffb41dd29-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13493-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cca289d2a4acd14c1cd9a84ffb41dd29-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=W6e384Lkjbw
|
https://papers.nips.cc/paper_files/paper/2021/file/cca289d2a4acd14c1cd9a84ffb41dd29-Supplemental.zip
|
We consider the problem of online linear regression in the stochastic setting. We derive high probability regret bounds for online $\textit{ridge}$ regression and the $\textit{forward}$ algorithm. This enables us to compare online regression algorithms more accurately and eliminate assumptions of bounded observations and predictions. Our study advocates for the use of the forward algorithm in lieu of ridge due to its enhanced bounds and robustness to the regularization parameter. Moreover, we explain how to integrate it in algorithms involving linear function approximation to remove a boundedness assumption without deteriorating theoretical bounds. We showcase this modification in linear bandit settings where it yields improved regret bounds. Last, we provide numerical experiments to illustrate our results and endorse our intuitions.
| null |
Dr Jekyll & Mr Hyde: the strange case of off-policy policy updates
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccb421d5f36c5a412816d494b15ca9f6-Abstract.html
|
Romain Laroche, Remi Tachet des Combes
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccb421d5f36c5a412816d494b15ca9f6-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13494-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ccb421d5f36c5a412816d494b15ca9f6-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=xVLzpMOexqo
|
https://papers.nips.cc/paper_files/paper/2021/file/ccb421d5f36c5a412816d494b15ca9f6-Supplemental.pdf
|
The policy gradient theorem states that the policy should only be updated in states that are visited by the current policy, which leads to insufficient planning in the off-policy states, and thus to convergence to suboptimal policies. We tackle this planning issue by extending the policy gradient theory to policy updates with respect to any state density. Under these generalized policy updates, we show convergence to optimality under a necessary and sufficient condition on the updates’ state densities, and thereby solve the aforementioned planning issue. We also prove asymptotic convergence rates that significantly improve those in the policy gradient literature. To implement the principles prescribed by our theory, we propose an agent, Dr Jekyll & Mr Hyde (J&H), with a double personality: Dr Jekyll purely exploits while Mr Hyde purely explores. J&H’s independent policies allow to record two separate replay buffers: one on-policy (Dr Jekyll’s) and one off-policy (Mr Hyde’s), and therefore to update J&H’s models with a mixture of on-policy and off-policy updates. More than an algorithm, J&H defines principles for actor-critic algorithms to satisfy the requirements we identify in our analysis. We extensively test on finite MDPs where J&H demonstrates a superior ability to recover from converging to a suboptimal policy without impairing its speed of convergence. We also implement a deep version of the algorithm and test it on a simple problem where it shows promising results.
| null |
Understanding Adaptive, Multiscale Temporal Integration In Deep Speech Recognition Systems
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccce2fab7336b8bc8362d115dec2d5a2-Abstract.html
|
Menoua Keshishian, Samuel Norman-Haignere, Nima Mesgarani
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccce2fab7336b8bc8362d115dec2d5a2-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13495-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ccce2fab7336b8bc8362d115dec2d5a2-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=h4es0CIohF
|
https://papers.nips.cc/paper_files/paper/2021/file/ccce2fab7336b8bc8362d115dec2d5a2-Supplemental.pdf
|
Natural signals such as speech are hierarchically structured across many different timescales, spanning tens (e.g., phonemes) to hundreds (e.g., words) of milliseconds, each of which is highly variable and context-dependent. While deep neural networks (DNNs) excel at recognizing complex patterns from natural signals, relatively little is known about how DNNs flexibly integrate across multiple timescales. Here, we show how a recently developed method for studying temporal integration in biological neural systems – the temporal context invariance (TCI) paradigm – can be used to understand temporal integration in DNNs. The method is simple: we measure responses to a large number of stimulus segments presented in two different contexts and estimate the smallest segment duration needed to achieve a context invariant response. We applied our method to understand how the popular DeepSpeech2 model learns to integrate across time in speech. We find that nearly all of the model units, even in recurrent layers, have a compact integration window within which stimuli substantially alter the response and outside of which stimuli have little effect. We show that training causes these integration windows to shrink at early layers and expand at higher layers, creating a hierarchy of integration windows across the network. Moreover, by measuring integration windows for time-stretched/compressed speech, we reveal a transition point, midway through the trained network, where integration windows become yoked to the duration of stimulus structures (e.g., phonemes or words) rather than absolute time. Similar phenomena were observed in a purely recurrent and purely convolutional network although structure-yoked integration was more prominent in the recurrent network. These findings suggest that deep speech recognition systems use a common motif to encode the hierarchical structure of speech: integrating across short, time-yoked windows at early layers and long, structure-yoked windows at later layers. Our method provides a straightforward and general-purpose toolkit for understanding temporal integration in black-box machine learning models.
| null |
VidLanKD: Improving Language Understanding via Video-Distilled Knowledge Transfer
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccdf3864e2fa9089f9eca4fc7a48ea0a-Abstract.html
|
Zineng Tang, Jaemin Cho, Hao Tan, Mohit Bansal
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccdf3864e2fa9089f9eca4fc7a48ea0a-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13496-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ccdf3864e2fa9089f9eca4fc7a48ea0a-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=nqSLT0WcZq
|
https://papers.nips.cc/paper_files/paper/2021/file/ccdf3864e2fa9089f9eca4fc7a48ea0a-Supplemental.pdf
|
Since visual perception can give rich information beyond text descriptions for world understanding, there has been increasing interest in leveraging visual grounding for language learning. Recently, vokenization (Tan and Bansal, 2020) has attracted attention by using the predictions of a text-to-image retrieval model as labels for language model supervision. Despite its success, the method suffers from approximation error of using finite image labels and the lack of vocabulary diversity of a small image-text dataset. To overcome these limitations, we present VidLanKD, a video-language knowledge distillation method for improving language understanding. We train a multi-modal teacher model on a video-text dataset, and then transfer its knowledge to a student language model with a text dataset. To avoid approximation error, we propose to use different knowledge distillation objectives. In addition, the use of a large-scale video-text dataset helps learn diverse and richer vocabularies. In our experiments, VidLanKD achieves consistent improvements over text-only language models and vokenization models, on several downstream language understanding tasks including GLUE, SQuAD, and SWAG. We also demonstrate the improved world knowledge, physical reasoning, and temporal reasoning capabilities of our model by evaluating on the GLUE-diagnostics, PIQA, and TRACIE datasets. Lastly, we present comprehensive ablation studies as well as visualizations of the learned text-to-video grounding results of our teacher and student language models.
| null |
Detecting Individual Decision-Making Style: Exploring Behavioral Stylometry in Chess
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccf8111910291ba472b385e9c5f59099-Abstract.html
|
Reid McIlroy-Young, Yu Wang, Siddhartha Sen, Jon Kleinberg, Ashton Anderson
|
https://papers.nips.cc/paper_files/paper/2021/hash/ccf8111910291ba472b385e9c5f59099-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13497-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ccf8111910291ba472b385e9c5f59099-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=9RFFgpQAOzk
|
https://papers.nips.cc/paper_files/paper/2021/file/ccf8111910291ba472b385e9c5f59099-Supplemental.pdf
|
The advent of machine learning models that surpass human decision-making ability in complex domains has initiated a movement towards building AI systems that interact with humans. Many building blocks are essential for this activity, with a central one being the algorithmic characterization of human behavior. While much of the existing work focuses on aggregate human behavior, an important long-range goal is to develop behavioral models that specialize to individual people and can differentiate among them.To formalize this process, we study the problem of behavioral stylometry, in which the task is to identify a decision-maker from their decisions alone. We present a transformer-based approach to behavioral stylometry in the context of chess, where one attempts to identify the player who played a set of games. Our method operates in a few-shot classification framework, and can correctly identify a player from among thousands of candidate players with 98% accuracy given only 100 labeled games. Even when trained on amateur play, our method generalises to out-of-distribution samples of Grandmaster players, despite the dramatic differences between amateur and world-class players. Finally, we consider more broadly what our resulting embeddings reveal about human style in chess, as well as the potential ethical implications of powerful methods for identifying individuals from behavioral data.
| null |
Coupled Gradient Estimators for Discrete Latent Variables
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd0b43eac0392accf3624b7372dec36e-Abstract.html
|
Zhe Dong, Andriy Mnih, George Tucker
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd0b43eac0392accf3624b7372dec36e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13498-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cd0b43eac0392accf3624b7372dec36e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=byizK1OI4xA
|
https://papers.nips.cc/paper_files/paper/2021/file/cd0b43eac0392accf3624b7372dec36e-Supplemental.pdf
|
Training models with discrete latent variables is challenging due to the high variance of unbiased gradient estimators. While low-variance reparameterization gradients of a continuous relaxation can provide an effective solution, a continuous relaxation is not always available or tractable. Dong et al. (2020) and Yin et al. (2020) introduced a performant estimator that does not rely on continuous relaxations; however, it is limited to binary random variables. We introduce a novel derivation of their estimator based on importance sampling and statistical couplings, which we extend to the categorical setting. Motivated by the construction of a stick-breaking coupling, we introduce gradient estimators based on reparameterizing categorical variables as sequences of binary variables and Rao-Blackwellization. In systematic experiments, we show that our proposed categorical gradient estimators provide state-of-the-art performance, whereas even with additional Rao-Blackwellization previous estimators (Yin et al., 2019) underperform a simpler REINFORCE with a leave-one-out-baseline estimator (Kool et al., 2019).
| null |
AutoGEL: An Automated Graph Neural Network with Explicit Link Information
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd3afef9b8b89558cd56638c3631868a-Abstract.html
|
Zhili Wang, Shimin DI, Lei Chen
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd3afef9b8b89558cd56638c3631868a-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13499-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cd3afef9b8b89558cd56638c3631868a-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=PftCCiHVQP
|
https://papers.nips.cc/paper_files/paper/2021/file/cd3afef9b8b89558cd56638c3631868a-Supplemental.pdf
|
Recently, Graph Neural Networks (GNNs) have gained popularity in a variety of real-world scenarios. Despite the great success, the architecture design of GNNs heavily relies on manual labor. Thus, automated graph neural network (AutoGNN) has attracted interest and attention from the research community, which makes significant performance improvements in recent years. However, existing AutoGNN works mainly adopt an implicit way to model and leverage the link information in the graphs, which is not well regularized to the link prediction task on graphs, and limits the performance of AutoGNN for other graph tasks. In this paper, we present a novel AutoGNN work that explicitly models the link information, abbreviated to AutoGEL. In such a way, AutoGEL can handle the link prediction task and improve the performance of AutoGNNs on the node classification and graph classification task. Moreover, AutoGEL proposes a novel search space containing various design dimensions at both intra-layer and inter-layer designs and adopts a more robust differentiable search algorithm to further improve efficiency and effectiveness. Experimental results on benchmark data sets demonstrate the superiority of AutoGEL on several tasks.
| null |
RL for Latent MDPs: Regret Guarantees and a Lower Bound
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd755a6c6b699f3262bcc2aa46ab507e-Abstract.html
|
Jeongyeol Kwon, Yonathan Efroni, Constantine Caramanis, Shie Mannor
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd755a6c6b699f3262bcc2aa46ab507e-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13500-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cd755a6c6b699f3262bcc2aa46ab507e-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=CLCVcl1rSPP
|
https://papers.nips.cc/paper_files/paper/2021/file/cd755a6c6b699f3262bcc2aa46ab507e-Supplemental.pdf
|
In this work, we consider the regret minimization problem for reinforcement learning in latent Markov Decision Processes (LMDP). In an LMDP, an MDP is randomly drawn from a set of $M$ possible MDPs at the beginning of the interaction, but the identity of the chosen MDP is not revealed to the agent. We first show that a general instance of LMDPs requires at least $\Omega((SA)^M)$ episodes to even approximate the optimal policy. Then, we consider sufficient assumptions under which learning good policies requires polynomial number of episodes. We show that the key link is a notion of separation between the MDP system dynamics. With sufficient separation, we provide an efficient algorithm with local guarantee, {\it i.e.,} providing a sublinear regret guarantee when we are given a good initialization. Finally, if we are given standard statistical sufficiency assumptions common in the Predictive State Representation (PSR) literature (e.g., \cite{boots2011online}) and a reachability assumption, we show that the need for initialization can be removed.
| null |
Adaptive Sampling for Minimax Fair Classification
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd7c230fc5deb01ff5f7b1be1acef9cf-Abstract.html
|
Shubhanshu Shekhar, Greg Fields, Mohammad Ghavamzadeh, Tara Javidi
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd7c230fc5deb01ff5f7b1be1acef9cf-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13501-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cd7c230fc5deb01ff5f7b1be1acef9cf-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=ZDMqRGSksHs
|
https://papers.nips.cc/paper_files/paper/2021/file/cd7c230fc5deb01ff5f7b1be1acef9cf-Supplemental.zip
|
Machine learning models trained on uncurated datasets can often end up adversely affecting inputs belonging to underrepresented groups. To address this issue, we consider the problem of adaptively constructing training sets which allow us to learn classifiers that are fair in a {\em minimax} sense. We first propose an adaptive sampling algorithm based on the principle of \emph{optimism}, and derive theoretical bounds on its performance. We also propose heuristic extensions of this algorithm suitable for application to large scale, practical problems. Next, by deriving algorithm independent lower-bounds for a specific class of problems, we show that the performance achieved by our adaptive scheme cannot be improved in general. We then validate the benefits of adaptively constructing training sets via experiments on synthetic tasks with logistic regression classifiers, as well as on several real-world tasks using convolutional neural networks (CNNs).
| null |
Structured in Space, Randomized in Time: Leveraging Dropout in RNNs for Efficient Training
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd81cfd0a3397761fac44ddbe5ec3349-Abstract.html
|
Anup Sarma, Sonali Singh, Huaipan Jiang, Rui Zhang, Mahmut Kandemir, Chita Das
|
https://papers.nips.cc/paper_files/paper/2021/hash/cd81cfd0a3397761fac44ddbe5ec3349-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13502-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cd81cfd0a3397761fac44ddbe5ec3349-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=m8KpGet0Etq
| null |
Recurrent Neural Networks (RNNs), more specifically their Long Short-Term Memory (LSTM) variants, have been widely used as a deep learning tool for tackling sequence-based learning tasks in text and speech. Training of such LSTM applications is computationally intensive due to the recurrent nature of hidden state computation that repeats for each time step. While sparsity in Deep Neural Nets has been widely seen as an opportunity for reducing computation time in both training and inference phases, the usage of non-ReLU activation in LSTM RNNs renders the opportunities for such dynamic sparsity associated with neuron activation and gradient values to be limited or non-existent. In this work, we identify dropout induced sparsity for LSTMs as a suitable mode of computation reduction. Dropout is a widely used regularization mechanism, which randomly drops computed neuron values during each iteration of training. We propose to structure dropout patterns, by dropping out the same set of physical neurons within a batch, resulting in column (row) level hidden state sparsity, which are well amenable to computation reduction at run-time in general-purpose SIMD hardware as well as systolic arrays. We provide a detailed analysis of how the dropout-induced sparsity propagates through the different stages of network training and how it can be leveraged in each stage. More importantly, our proposed approach works as a direct replacement for existing dropout-based application settings. We conduct our experiments for three representative NLP tasks: language modelling on the PTB dataset, OpenNMT based machine translation using the IWSLT De-En and En-Vi datasets, and named entity recognition sequence labelling using the CoNLL-2003 shared task. We demonstrate that our proposed approach can be used to translate dropout-based computation reduction into reduced training time, with improvement ranging from 1.23$\times$ to 1.64$\times$, without sacrificing the target metric.
| null |
Variational Continual Bayesian Meta-Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdd0500dc0ef6682fa6ec6d2e6b577c4-Abstract.html
|
Qiang Zhang, Jinyuan Fang, Zaiqiao Meng, Shangsong Liang, Emine Yilmaz
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdd0500dc0ef6682fa6ec6d2e6b577c4-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13503-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cdd0500dc0ef6682fa6ec6d2e6b577c4-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=VH2og5jlrzm
|
https://papers.nips.cc/paper_files/paper/2021/file/cdd0500dc0ef6682fa6ec6d2e6b577c4-Supplemental.pdf
|
Conventional meta-learning considers a set of tasks from a stationary distribution. In contrast, this paper focuses on a more complex online setting, where tasks arrive sequentially and follow a non-stationary distribution. Accordingly, we propose a Variational Continual Bayesian Meta-Learning (VC-BML) algorithm. VC-BML maintains a Dynamic Gaussian Mixture Model for meta-parameters, with the number of component distributions determined by a Chinese Restaurant Process. Dynamic mixtures at the meta-parameter level increase the capability to adapt to diverse tasks due to a larger parameter space, alleviating the negative knowledge transfer problem. To infer posteriors of model parameters, compared to the previously used point estimation method, we develop a more robust posterior approximation method -- structured variational inference for the sake of avoiding forgetting knowledge. Experiments on tasks from non-stationary distributions show that VC-BML is superior in transferring knowledge among diverse tasks and alleviating catastrophic forgetting in an online setting.
| null |
Recognizing Vector Graphics without Rasterization
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdf1035c34ec380218a8cc9a43d438f9-Abstract.html
|
XINYANG JIANG, LU LIU, Caihua Shan, Yifei Shen, Xuanyi Dong, Dongsheng Li
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdf1035c34ec380218a8cc9a43d438f9-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13504-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cdf1035c34ec380218a8cc9a43d438f9-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=_ZXlOpdufFJ
|
https://papers.nips.cc/paper_files/paper/2021/file/cdf1035c34ec380218a8cc9a43d438f9-Supplemental.pdf
|
In this paper, we consider a different data format for images: vector graphics. In contrast to raster graphics which are widely used in image recognition, vector graphics can be scaled up or down into any resolution without aliasing or information loss, due to the analytic representation of the primitives in the document. Furthermore, vector graphics are able to give extra structural information on how low-level elements group together to form high level shapes or structures. These merits of graphic vectors have not been fully leveraged in existing methods. To explore this data format, we target on the fundamental recognition tasks: object localization and classification. We propose an efficient CNN-free pipeline that does not render the graphic into pixels (i.e. rasterization), and takes textual document of the vector graphics as input, called YOLaT (You Only Look at Text). YOLaT builds multi-graphs to model the structural and spatial information in vector graphics, and a dual-stream graph neural network is proposed to detect objects from the graph. Our experiments show that by directly operating on vector graphics, YOLaT outperforms raster-graphic based object detection baselines in terms of both average precision and efficiency. Code is available at https://github.com/microsoft/YOLaT-VectorGraphicsRecognition.
| null |
On Episodes, Prototypical Networks, and Few-Shot Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdfa4c42f465a5a66871587c69fcfa34-Abstract.html
|
Steinar Laenen, Luca Bertinetto
|
https://papers.nips.cc/paper_files/paper/2021/hash/cdfa4c42f465a5a66871587c69fcfa34-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13505-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cdfa4c42f465a5a66871587c69fcfa34-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=bJaZ8leI0QJ
|
https://papers.nips.cc/paper_files/paper/2021/file/cdfa4c42f465a5a66871587c69fcfa34-Supplemental.pdf
|
Episodic learning is a popular practice among researchers and practitioners interested in few-shot learning.It consists of organising training in a series of learning problems (or episodes), each divided into a small training and validation subset to mimic the circumstances encountered during evaluation.But is this always necessary?In this paper, we investigate the usefulness of episodic learning in methods which use nonparametric approaches, such as nearest neighbours, at the level of the episode.For these methods, we not only show how the constraints imposed by episodic learning are not necessary, but that they in fact lead to a data-inefficient way of exploiting training batches.We conduct a wide range of ablative experiments with Matching and Prototypical Networks, two of the most popular methods that use nonparametric approaches at the level of the episode.Their "non-episodic'' counterparts are considerably simpler, have less hyperparameters, and improve their performance in multiple few-shot classification datasets.
| null |
Pointwise Bounds for Distribution Estimation under Communication Constraints
|
https://papers.nips.cc/paper_files/paper/2021/hash/ce4449660c6523b377b22a1dc2da5556-Abstract.html
|
Wei-Ning Chen, Peter Kairouz, Ayfer Ozgur
|
https://papers.nips.cc/paper_files/paper/2021/hash/ce4449660c6523b377b22a1dc2da5556-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13506-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ce4449660c6523b377b22a1dc2da5556-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=l41jc6kUfKr
|
https://papers.nips.cc/paper_files/paper/2021/file/ce4449660c6523b377b22a1dc2da5556-Supplemental.pdf
|
We consider the problem of estimating a $d$-dimensional discrete distribution from its samples observed under a $b$-bit communication constraint. In contrast to most previous results that largely focus on the global minimax error, we study the local behavior of the estimation error and provide \emph{pointwise} bounds that depend on the target distribution $p$. In particular, we show that the $\ell_2$ error decays with $O\left(\frac{\lVert p\rVert_{1/2}}{n2^b}\vee \frac{1}{n}\right)$ when $n$ is sufficiently large, hence it is governed by the \emph{half-norm} of $p$ instead of the ambient dimension $d$. For the achievability result, we propose a two-round sequentially interactive estimation scheme that achieves this error rate uniformly over all $p$. Our scheme is based on a novel local refinement idea, where we first use a standard global minimax scheme to localize $p$ and then use the remaining samples to locally refine our estimate.We also develop a new local minimax lower bound with (almost) matching $\ell_2$ error, showing that any interactive scheme must admit a $\Omega\left( \frac{\lVert p \rVert_{{(1+\delta)}/{2}}}{n2^b}\right)$ $\ell_2$ error for any $\delta > 0$ when $n$ is sufficiently large. The lower bound is derived by first finding the best parametric sub-model containing $p$, and then upper bounding the quantized Fisher information under this model. Our upper and lower bounds together indicate that the $\mathsf{H}_{1/2}(p) = \log(\lVert p \rVert_{{1}/{2}})$ bits of communication is both sufficient and necessary to achieve the optimal (centralized) performance, where $\mathsf{H}_{{1}/{2}}(p)$ is the R\'enyi entropy of order $2$. Therefore, under the $\ell_2$ loss, the correct measure of the local communication complexity at $p$ is its R\'enyi entropy.
| null |
CHIP: CHannel Independence-based Pruning for Compact Neural Networks
|
https://papers.nips.cc/paper_files/paper/2021/hash/ce6babd060aa46c61a5777902cca78af-Abstract.html
|
Yang Sui, Miao Yin, Yi Xie, Huy Phan, Saman Aliari Zonouz, Bo Yuan
|
https://papers.nips.cc/paper_files/paper/2021/hash/ce6babd060aa46c61a5777902cca78af-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13507-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ce6babd060aa46c61a5777902cca78af-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=EmeWbcWORRg
|
https://papers.nips.cc/paper_files/paper/2021/file/ce6babd060aa46c61a5777902cca78af-Supplemental.pdf
|
Filter pruning has been widely used for neural network compression because of its enabled practical acceleration. To date, most of the existing filter pruning works explore the importance of filters via using intra-channel information. In this paper, starting from an inter-channel perspective, we propose to perform efficient filter pruning using Channel Independence, a metric that measures the correlations among different feature maps. The less independent feature map is interpreted as containing less useful information$/$knowledge, and hence its corresponding filter can be pruned without affecting model capacity. We systematically investigate the quantification metric, measuring scheme and sensitiveness$/$reliability of channel independence in the context of filter pruning. Our evaluation results for different models on various datasets show the superior performance of our approach. Notably, on CIFAR-10 dataset our solution can bring $0.75\%$ and $0.94\%$ accuracy increase over baseline ResNet-56 and ResNet-110 models, respectively, and meanwhile the model size and FLOPs are reduced by $42.8\%$ and $47.4\%$ (for ResNet-56) and $48.3\%$ and $52.1\%$ (for ResNet-110), respectively. On ImageNet dataset, our approach can achieve $40.8\%$ and $44.8\%$ storage and computation reductions, respectively, with $0.15\%$ accuracy increase over the baseline ResNet-50 model. The code is available at https://github.com/Eclipsess/CHIP_NeurIPS2021.
| null |
Federated Split Task-Agnostic Vision Transformer for COVID-19 CXR Diagnosis
|
https://papers.nips.cc/paper_files/paper/2021/hash/ceb0595112db2513b9325a85761b7310-Abstract.html
|
Sangjoon Park, Gwanghyun Kim, Jeongsol Kim, Boah Kim, Jong Chul Ye
|
https://papers.nips.cc/paper_files/paper/2021/hash/ceb0595112db2513b9325a85761b7310-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13508-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/ceb0595112db2513b9325a85761b7310-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=Ggikq6Tdxch
|
https://papers.nips.cc/paper_files/paper/2021/file/ceb0595112db2513b9325a85761b7310-Supplemental.pdf
|
Federated learning, which shares the weights of the neural network across clients, is gaining attention in the healthcare sector as it enables training on a large corpus of decentralized data while maintaining data privacy. For example, this enables neural network training for COVID-19 diagnosis on chest X-ray (CXR) images without collecting patient CXR data across multiple hospitals. Unfortunately, the exchange of the weights quickly consumes the network bandwidth if highly expressive network architecture is employed. So-called split learning partially solves this problem by dividing a neural network into a client and a server part, so that the client part of the network takes up less extensive computation resources and bandwidth. However, it is not clear how to find the optimal split without sacrificing the overall network performance. To amalgamate these methods and thereby maximize their distinct strengths, here we show that the Vision Transformer, a recently developed deep learning architecture with straightforward decomposable configuration, is ideally suitable for split learning without sacrificing performance. Even under the non-independent and identically distributed data distribution which emulates a real collaboration between hospitals using CXR datasets from multiple sources, the proposed framework was able to attain performance comparable to data-centralized training. In addition, the proposed framework along with heterogeneous multi-task clients also improves individual task performances including the diagnosis of COVID-19, eliminating the need for sharing large weights with innumerable parameters. Our results affirm the suitability of Transformer for collaborative learning in medical imaging and pave the way forward for future real-world implementations.
| null |
Active Offline Policy Selection
|
https://papers.nips.cc/paper_files/paper/2021/hash/cec2346566ba8ecd04bfd992fd193fb3-Abstract.html
|
Ksenia Konyushova, Yutian Chen, Thomas Paine, Caglar Gulcehre, Cosmin Paduraru, Daniel J. Mankowitz, Misha Denil, Nando de Freitas
|
https://papers.nips.cc/paper_files/paper/2021/hash/cec2346566ba8ecd04bfd992fd193fb3-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13509-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cec2346566ba8ecd04bfd992fd193fb3-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=Zsrn9wXWN0
|
https://papers.nips.cc/paper_files/paper/2021/file/cec2346566ba8ecd04bfd992fd193fb3-Supplemental.pdf
|
This paper addresses the problem of policy selection in domains with abundant logged data, but with a restricted interaction budget. Solving this problem would enable safe evaluation and deployment of offline reinforcement learning policies in industry, robotics, and recommendation domains among others. Several off-policy evaluation (OPE) techniques have been proposed to assess the value of policies using only logged data. However, there is still a big gap between the evaluation by OPE and the full online evaluation in the real environment. Yet, large amounts of online interactions are often not possible in practice. To overcome this problem, we introduce active offline policy selection --- a novel sequential decision approach that combines logged data with online interaction to identify the best policy. This approach uses OPE estimates to warm start the online evaluation. Then, in order to utilize the limited environment interactions wisely we decide which policy to evaluate next based on a Bayesian optimization method with a kernel function that represents policy similarity. We use multiple benchmarks with a large number of candidate policies to show that the proposed approach improves upon state-of-the-art OPE estimates and pure online policy evaluation.
| null |
Unsupervised Representation Transfer for Small Networks: I Believe I Can Distill On-the-Fly
|
https://papers.nips.cc/paper_files/paper/2021/hash/cecd845e3577efdaaf24eea03af4c033-Abstract.html
|
Hee Min Choi, Hyoa Kang, Dokwan Oh
|
https://papers.nips.cc/paper_files/paper/2021/hash/cecd845e3577efdaaf24eea03af4c033-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13510-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cecd845e3577efdaaf24eea03af4c033-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=BYrJYl1rexa
|
https://papers.nips.cc/paper_files/paper/2021/file/cecd845e3577efdaaf24eea03af4c033-Supplemental.pdf
|
A current remarkable improvement of unsupervised visual representation learning is based on heavy networks with large-batch training. While recent methods have greatly reduced the gap between supervised and unsupervised performance of deep models such as ResNet-50, this development has been relatively limited for small models. In this work, we propose a novel unsupervised learning framework for small networks that combines deep self-supervised representation learning and knowledge distillation within one-phase training. In particular, a teacher model is trained to produce consistent cluster assignments between different views of the same image. Simultaneously, a student model is encouraged to mimic the prediction of on-the-fly self-supervised teacher. For effective knowledge transfer, we adopt the idea of domain classifier so that student training is guided by discriminative features invariant to the representational space shift between teacher and student. We also introduce a network driven multi-view generation paradigm to capture rich feature information contained in the network itself. Extensive experiments show that our student models surpass state-of-the-art offline distilled networks even from stronger self-supervised teachers as well as top-performing self-supervised models. Notably, our ResNet-18, trained with ResNet-50 teacher, achieves 68.3% ImageNet Top-1 accuracy on frozen feature linear evaluation, which is only 1.5% below the supervised baseline.
| null |
Understanding Bandits with Graph Feedback
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf004fdc76fa1a4f25f62e0eb5261ca3-Abstract.html
|
Houshuang Chen, zengfeng Huang, Shuai Li, Chihao Zhang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf004fdc76fa1a4f25f62e0eb5261ca3-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13511-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf004fdc76fa1a4f25f62e0eb5261ca3-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=V3aZTKsHykQ
| null |
The bandit problem with graph feedback, proposed in [Mannor and Shamir, NeurIPS 2011], is modeled by a directed graph $G=(V,E)$ where $V$ is the collection of bandit arms, and once an arm is triggered, all its incident arms are observed. A fundamental question is how the structure of the graph affects the min-max regret. We propose the notions of the fractional weak domination number $\delta^*$ and the $k$-packing independence number capturing upper bound and lower bound for the regret respectively. We show that the two notions are inherently connected via aligning them with the linear program of the weakly dominating set and its dual --- the fractional vertex packing set respectively. Based on this connection, we utilize the strong duality theorem to prove a general regret upper bound $O\left(\left(\delta^*\log |V|\right)^{\frac{1}{3}}T^{\frac{2}{3}}\right)$ and a lower bound $\Omega\left(\left(\delta^*/\alpha\right)^{\frac{1}{3}}T^{\frac{2}{3}}\right)$ where $\alpha$ is the integrality gap of the dual linear program. Therefore, our bounds are tight up to a $\left(\log |V|\right)^{\frac{1}{3}}$ factor on graphs with bounded integrality gap for the vertex packing problem including trees and graphs with bounded degree. Moreover, we show that for several special families of graphs, we can get rid of the $\left(\log |V|\right)^{\frac{1}{3}}$ factor and establish optimal regret.
| null |
Information-theoretic generalization bounds for black-box learning algorithms
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf0d02ec99e61a64137b8a2c3b03e030-Abstract.html
|
Hrayr Harutyunyan, Maxim Raginsky, Greg Ver Steeg, Aram Galstyan
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf0d02ec99e61a64137b8a2c3b03e030-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13512-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf0d02ec99e61a64137b8a2c3b03e030-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=L_cN8vD0XdT
|
https://papers.nips.cc/paper_files/paper/2021/file/cf0d02ec99e61a64137b8a2c3b03e030-Supplemental.pdf
|
We derive information-theoretic generalization bounds for supervised learning algorithms based on the information contained in predictions rather than in the output of the training algorithm. These bounds improve over the existing information-theoretic bounds, are applicable to a wider range of algorithms, and solve two key challenges: (a) they give meaningful results for deterministic algorithms and (b) they are significantly easier to estimate. We show experimentally that the proposed bounds closely follow the generalization gap in practical scenarios for deep learning.
| null |
Trash or Treasure? An Interactive Dual-Stream Strategy for Single Image Reflection Separation
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf1f78fe923afe05f7597da2be7a3da8-Abstract.html
|
Qiming Hu, Xiaojie Guo
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf1f78fe923afe05f7597da2be7a3da8-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13513-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf1f78fe923afe05f7597da2be7a3da8-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=3Ky3sH5enrc
| null |
Single image reflection separation (SIRS), as a representative blind source separation task, aims to recover two layers, $\textit{i.e.}$, transmission and reflection, from one mixed observation, which is challenging due to the highly ill-posed nature. Existing deep learning based solutions typically restore the target layers individually, or with some concerns at the end of the output, barely taking into account the interaction across the two streams/branches. In order to utilize information more efficiently, this work presents a general yet simple interactive strategy, namely $\textit{your trash is my treasure}$ (YTMT), for constructing dual-stream decomposition networks. To be specific, we explicitly enforce the two streams to communicate with each other block-wisely. Inspired by the additive property between the two components, the interactive path can be easily built via transferring, instead of discarding, deactivated information by the ReLU rectifier from one stream to the other. Both ablation studies and experimental results on widely-used SIRS datasets are conducted to demonstrate the efficacy of YTMT, and reveal its superiority over other state-of-the-art alternatives. The implementation is quite simple and our code is publicly available at https://github.com/mingcv/YTMT-Strategy.
| null |
Rot-Pro: Modeling Transitivity by Projection in Knowledge Graph Embedding
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf2f3fe19ffba462831d7f037a07fc83-Abstract.html
|
Tengwei Song, Jie Luo, Lei Huang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf2f3fe19ffba462831d7f037a07fc83-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13514-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf2f3fe19ffba462831d7f037a07fc83-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=P4W74BXoyBy
|
https://papers.nips.cc/paper_files/paper/2021/file/cf2f3fe19ffba462831d7f037a07fc83-Supplemental.pdf
|
Knowledge graph embedding models learn the representations of entities and relations in the knowledge graphs for predicting missing links (relations) between entities. Their effectiveness are deeply affected by the ability of modeling and inferring different relation patterns such as symmetry, asymmetry, inversion, composition and transitivity. Although existing models are already able to model many of these relations patterns, transitivity, a very common relation pattern, is still not been fully supported. In this paper, we first theoretically show that the transitive relations can be modeled with projections. We then propose the Rot-Pro model which combines the projection and relational rotation together. We prove that Rot-Pro can infer all the above relation patterns. Experimental results show that the proposed Rot-Pro model effectively learns the transitivity pattern and achieves the state-of-the-art results on the link prediction task in the datasets containing transitive relations.
| null |
Planning from Pixels in Environments with Combinatorially Hard Search Spaces
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf708fc1decf0337aded484f8f4519ae-Abstract.html
|
Marco Bagatella, Miroslav Olšák, Michal Rolínek, Georg Martius
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf708fc1decf0337aded484f8f4519ae-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13515-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf708fc1decf0337aded484f8f4519ae-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=XgGUUaKgips
|
https://papers.nips.cc/paper_files/paper/2021/file/cf708fc1decf0337aded484f8f4519ae-Supplemental.pdf
|
The ability to form complex plans based on raw visual input is a litmus test for current capabilities of artificial intelligence, as it requires a seamless combination of visual processing and abstract algorithmic execution, two traditionally separate areas of computer science. A recent surge of interest in this field brought advances that yield good performance in tasks ranging from arcade games to continuous control; these methods however do not come without significant issues, such as limited generalization capabilities and difficulties when dealing with combinatorially hard planning instances. Our contribution is two-fold: (i) we present a method that learns to represent its environment as a latent graph and leverages state reidentification to reduce the complexity of finding a good policy from exponential to linear (ii) we introduce a set of lightweight environments with an underlying discrete combinatorial structure in which planning is challenging even for humans. Moreover, we show that our methods achieves strong empirical generalization to variations in the environment, even across highly disadvantaged regimes, such as “one-shot” planning, or in an offline RL paradigm which only provides low-quality trajectories.
| null |
PLUGIn: A simple algorithm for inverting generative models with recovery guarantees
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf77e1f8490495e9f8dedceaf372f969-Abstract.html
|
Babhru Joshi, Xiaowei Li, Yaniv Plan, Ozgur Yilmaz
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf77e1f8490495e9f8dedceaf372f969-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13516-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf77e1f8490495e9f8dedceaf372f969-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=rxAS126OC-A
|
https://papers.nips.cc/paper_files/paper/2021/file/cf77e1f8490495e9f8dedceaf372f969-Supplemental.pdf
|
We consider the problem of recovering an unknown latent code vector under a known generative model. For a $d$-layer deep generative network $\mathcal{G}:\mathbb{R}^{n_0}\rightarrow \mathbb{R}^{n_d}$ with ReLU activation functions, let the observation be $\mathcal{G}(x)+\epsilon$ where $\epsilon$ is noise. We introduce a simple novel algorithm, Partially Linearized Update for Generative Inversion (PLUGIn), to estimate $x$ (and thus $\mathcal{G}(x)$). We prove that, when weights are Gaussian and layer widths $n_i \gtrsim 5^i n_0$ (up to log factors), the algorithm converges geometrically to a neighbourhood of $x$ with high probability. Note the inequality on layer widths allows $n_i>n_{i+1}$ when $i\geq 1$. To our knowledge, this is the first such result for networks with some contractive layers. After a sufficient number of iterations, the estimation errors for both $x$ and $\mathcal{G}(x)$ are at most in the order of $\sqrt{4^dn_0/n_d} \|\epsilon\|$. Thus, the algorithm can denoise when the expansion ratio $n_d/n_0$ is large. Numerical experiments on synthetic data and real data are provided to validate our theoretical results and to illustrate that the algorithm can effectively remove artifacts in an image.
| null |
Modular Gaussian Processes for Transfer Learning
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf79ae6addba60ad018347359bd144d2-Abstract.html
|
Pablo Moreno-Muñoz, Antonio Artes, Mauricio Álvarez
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf79ae6addba60ad018347359bd144d2-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13517-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf79ae6addba60ad018347359bd144d2-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=7U7JxTiL8gz
|
https://papers.nips.cc/paper_files/paper/2021/file/cf79ae6addba60ad018347359bd144d2-Supplemental.pdf
|
We present a framework for transfer learning based on modular variational Gaussian processes (GP). We develop a module-based method that having a dictionary of well fitted GPs, each model being characterised by its hyperparameters, pseudo-inputs and their corresponding posterior densities, one could build ensemble GP models without revisiting any data. Our method avoids undesired data centralisation, reduces rising computational costs and allows the transfer of learned uncertainty metrics after training. We exploit the augmentation of high-dimensional integral operators based on the Kullback-Leibler divergence between stochastic processes to introduce an efficient lower bound under all the sparse variational GPs, with different complexity and even likelihood distribution. The method is also valid for multi-output GPs, learning correlations a posteriori between independent modules. Extensive results illustrate the usability of our framework in large-scale and multi-task experiments, also compared with the exact inference methods in the literature.
| null |
Neural Human Performer: Learning Generalizable Radiance Fields for Human Performance Rendering
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf866614b6b18cda13fe699a3a65661b-Abstract.html
|
Youngjoong Kwon, Dahun Kim, Duygu Ceylan, Henry Fuchs
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf866614b6b18cda13fe699a3a65661b-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13518-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf866614b6b18cda13fe699a3a65661b-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=YFysbLCFdIe
|
https://papers.nips.cc/paper_files/paper/2021/file/cf866614b6b18cda13fe699a3a65661b-Supplemental.pdf
|
In this paper, we aim at synthesizing a free-viewpoint video of an arbitrary human performance using sparse multi-view cameras. Recently, several works have addressed this problem by learning person-specific neural radiance fields (NeRF) to capture the appearance of a particular human. In parallel, some work proposed to use pixel-aligned features to generalize radiance fields to arbitrary new scenes and objects. Adopting such generalization approaches to humans, however, is highly challenging due to the heavy occlusions and dynamic articulations of body parts. To tackle this, we propose Neural Human Performer, a novel approach that learns generalizable neural radiance fields based on a parametric human body model for robust performance capture. Specifically, we first introduce a temporal transformer that aggregates tracked visual features based on the skeletal body motion over time. Moreover, a multi-view transformer is proposed to perform cross-attention between the temporally-fused features and the pixel-aligned features at each time step to integrate observations on the fly from multiple views. Experiments on the ZJU-MoCap and AIST datasets show that our method significantly outperforms recent generalizable NeRF methods on unseen identities and poses.
| null |
Locally differentially private estimation of functionals of discrete distributions
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf8c9be2a4508a24ae92c9d3d379131d-Abstract.html
|
Cristina Butucea, Yann ISSARTEL
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf8c9be2a4508a24ae92c9d3d379131d-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13519-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf8c9be2a4508a24ae92c9d3d379131d-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=cjnSJIf3c9Y
| null |
We study the problem of estimating non-linear functionals of discrete distributions in the context of local differential privacy. The initial data $x_1,\ldots,x_n \in[K]$ are supposed i.i.d. and distributed according to an unknown discrete distribution $p = (p_1,\ldots,p_K)$. Only $\alpha$-locally differentially private (LDP) samples $z_1,...,z_n$ are publicly available, where the term 'local' means that each $z_i$ is produced using one individual attribute $x_i$. We exhibit privacy mechanisms (PM) that are interactive (i.e. they are allowed to use already published confidential data) or non-interactive. We describe the behavior of the quadratic risk for estimating the power sum functional $F_{\gamma} = \sum_{k=1}^K p_k^{\gamma}$, $\gamma >0$ as a function of $K, \, n$ and $\alpha$. In the non-interactive case, we study twol plug-in type estimators of $F_{\gamma}$, for all $\gamma >0$, that are similar to the MLE analyzed by Jiao et al. (2017) in the multinomial model. However, due to the privacy constraint the rates we attain are slower and similar to those obtained in the Gaussian model by Collier et al. (2020). In the sequentially interactive case, we introduce for all $\gamma >1$ a two-step procedure which attains the parametric rate $(n \alpha^2)^{-1/2}$ when $\gamma \geq 2$. We give lower bounds results over all $\alpha-$LDP mechanisms and over all estimators using the private samples.
| null |
Asymptotics of representation learning in finite Bayesian neural networks
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf9dc5e4e194fc21f397b4cac9cc3ae9-Abstract.html
|
Jacob Zavatone-Veth, Abdulkadir Canatar, Ben Ruben, Cengiz Pehlevan
|
https://papers.nips.cc/paper_files/paper/2021/hash/cf9dc5e4e194fc21f397b4cac9cc3ae9-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13520-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cf9dc5e4e194fc21f397b4cac9cc3ae9-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=1oRFmD0Fl-5
|
https://papers.nips.cc/paper_files/paper/2021/file/cf9dc5e4e194fc21f397b4cac9cc3ae9-Supplemental.pdf
|
Recent works have suggested that finite Bayesian neural networks may sometimes outperform their infinite cousins because finite networks can flexibly adapt their internal representations. However, our theoretical understanding of how the learned hidden layer representations of finite networks differ from the fixed representations of infinite networks remains incomplete. Perturbative finite-width corrections to the network prior and posterior have been studied, but the asymptotics of learned features have not been fully characterized. Here, we argue that the leading finite-width corrections to the average feature kernels for any Bayesian network with linear readout and Gaussian likelihood have a largely universal form. We illustrate this explicitly for three tractable network architectures: deep linear fully-connected and convolutional networks, and networks with a single nonlinear hidden layer. Our results begin to elucidate how task-relevant learning signals shape the hidden layer representations of wide Bayesian neural networks.
| null |
Adaptive Ensemble Q-learning: Minimizing Estimation Bias via Error Feedback
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfa45151ccad6bf11ea146ed563f2119-Abstract.html
|
Hang Wang, Sen Lin, Junshan Zhang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfa45151ccad6bf11ea146ed563f2119-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13521-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cfa45151ccad6bf11ea146ed563f2119-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=YL6e9oSeInj
|
https://papers.nips.cc/paper_files/paper/2021/file/cfa45151ccad6bf11ea146ed563f2119-Supplemental.pdf
|
The ensemble method is a promising way to mitigate the overestimation issue in Q-learning, where multiple function approximators are used to estimate the action values. It is known that the estimation bias hinges heavily on the ensemble size (i.e., the number of Q-function approximators used in the target), and that determining the 'right' ensemble size is highly nontrivial, because of the time-varying nature of the function approximation errors during the learning process. To tackle this challenge, we first derive an upper bound and a lower bound on the estimation bias, based on which the ensemble size is adapted to drive the bias to be nearly zero, thereby coping with the impact of the time-varying approximation errors accordingly. Motivated by the theoretic findings, we advocate that the ensemble method can be combined with Model Identification Adaptive Control (MIAC) for effective ensemble size adaptation. Specifically, we devise Adaptive Ensemble Q-learning (AdaEQ), a generalized ensemble method with two key steps: (a) approximation error characterization which serves as the feedback for flexibly controlling the ensemble size, and (b) ensemble size adaptation tailored towards minimizing the estimation bias. Extensive experiments are carried out to show that AdaEQ can improve the learning performance than the existing methods for the MuJoCo benchmark.
| null |
Domain Adaptation with Invariant Representation Learning: What Transformations to Learn?
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfc5d9422f0c8f8ad796711102dbe32b-Abstract.html
|
Petar Stojanov, Zijian Li, Mingming Gong, Ruichu Cai, Jaime Carbonell, Kun Zhang
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfc5d9422f0c8f8ad796711102dbe32b-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13522-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cfc5d9422f0c8f8ad796711102dbe32b-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=zdmF437BCB
|
https://papers.nips.cc/paper_files/paper/2021/file/cfc5d9422f0c8f8ad796711102dbe32b-Supplemental.pdf
|
Unsupervised domain adaptation, as a prevalent transfer learning setting, spans many real-world applications. With the increasing representational power and applicability of neural networks, state-of-the-art domain adaptation methods make use of deep architectures to map the input features $X$ to a latent representation $Z$ that has the same marginal distribution across domains. This has been shown to be insufficient for generating optimal representation for classification, and to find conditionally invariant representations, usually strong assumptions are needed. We provide reasoning why when the supports of the source and target data from overlap, any map of $X$ that is fixed across domains may not be suitable for domain adaptation via invariant features. Furthermore, we develop an efficient technique in which the optimal map from $X$ to $Z$ also takes domain-specific information as input, in addition to the features $X$. By using the property of minimal changes of causal mechanisms across domains, our model also takes into account the domain-specific information to ensure that the latent representation $Z$ does not discard valuable information about $Y$. We demonstrate the efficacy of our method via synthetic and real-world data experiments. The code is available at: \texttt{https://github.com/DMIRLAB-Group/DSAN}.
| null |
CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfe8504bda37b575c70ee1a8276f3486-Abstract.html
|
Yusuke Tashiro, Jiaming Song, Yang Song, Stefano Ermon
|
https://papers.nips.cc/paper_files/paper/2021/hash/cfe8504bda37b575c70ee1a8276f3486-Abstract.html
|
NIPS 2021
|
https://papers.nips.cc/paper_files/paper/13523-/bibtex
|
https://papers.nips.cc/paper_files/paper/2021/file/cfe8504bda37b575c70ee1a8276f3486-Paper.pdf
|
https://papers.nips.cchttps://openreview.net/forum?id=VzuIzbRDrum
|
https://papers.nips.cc/paper_files/paper/2021/file/cfe8504bda37b575c70ee1a8276f3486-Supplemental.pdf
|
The imputation of missing values in time series has many applications in healthcare and finance. While autoregressive models are natural candidates for time series imputation, score-based diffusion models have recently outperformed existing counterparts including autoregressive models in many tasks such as image generation and audio synthesis, and would be promising for time series imputation. In this paper, we propose Conditional Score-based Diffusion model (CSDI), a novel time series imputation method that utilizes score-based diffusion models conditioned on observed data. Unlike existing score-based approaches, the conditional diffusion model is explicitly trained for imputation and can exploit correlations between observed values. On healthcare and environmental data, CSDI improves by 40-65% over existing probabilistic imputation methods on popular performance metrics. In addition, deterministic imputation by CSDI reduces the error by 5-20% compared to the state-of-the-art deterministic imputation methods. Furthermore, CSDI can also be applied to time series interpolation and probabilistic forecasting, and is competitive with existing baselines. The code is available at https://github.com/ermongroup/CSDI.
| null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.