title
stringlengths 7
246
| abstract
stringlengths 3
3.31k
|
|---|---|
Multi-level Binarized LSTM in EEG Classification for Wearable Devices | Long Short-Term Memory (LSTM) is widely used in various sequential applications. Complex LSTMs could be hardly deployed on wearable and resourced-limited devices due to the huge amount of computations and memory requirements. Binary LSTMs are introduced to cope with this problem, however, they lead to significant accuracy loss in some application such as EEG classification which is essential to be deployed in wearable devices. In this paper, we propose an efficient multi-level binarized LSTM which has significantly reduced computations whereas ensuring an accuracy pretty close to full precision LSTM. By deploying 5-level binarized weights and inputs, our method reduces area and delay of MAC operation about 31* and 27* in 65nm technology, respectively with less than 0.01% accuracy loss. In contrast to many compute-intensive deep-learning approaches, the proposed algorithm is lightweight, and therefore, brings performance efficiency with accurate LSTM-based EEG classification to real-time wearable devices. |
Bayes Optimal Early Stopping Policies for Black-Box Optimization | We derive an optimal policy for adaptively restarting a randomized algorithm, based on observed features of the run-so-far, so as to minimize the expected time required for the algorithm to successfully terminate. Given a suitable Bayesian prior, this result can be used to select the optimal black-box optimization algorithm from among a large family of algorithms that includes random search, Successive Halving, and Hyperband. On CIFAR-10 and ImageNet hyperparameter tuning problems, the proposed policies offer up to a factor of 13 improvement over random search in terms of expected time to reach a given target accuracy, and up to a factor of 3 improvement over a baseline adaptive policy that terminates a run whenever its accuracy is below-median. |
TOMA: Topological Map Abstraction for Reinforcement Learning | Animals are able to discover the topological map (graph) of surrounding environment, which will be used for navigation. Inspired by this biological phenomenon, researchers have recently proposed to generate graph representation for Markov decision process (MDP) and use such graphs for planning in reinforcement learning (RL). However, existing graph generation methods suffer from many drawbacks. One drawback is that existing methods do not learn an abstraction for graphs, which results in high memory and computation cost. This drawback also makes generated graph non-robust, which degrades the planning performance. Another drawback is that existing methods cannot be used for facilitating exploration which is important in RL. In this paper, we propose a new method, called topological map abstraction (TOMA), for graph generation. TOMA can generate an abstract graph representation for MDP, which costs much less memory and computation cost than existing methods. Furthermore, TOMA can be used for facilitating exploration. In particular, we propose planning to explore, in which TOMA is used to accelerate exploration by guiding the agent towards unexplored states. A novel experience replay module called vertex memory is also proposed to improve exploration performance. Experimental results show that TOMA can outperform existing methods to achieve the state-of-the-art performance. |
Causal Inference with Corrupted Data: Measurement Error, Missing Values, Discretization, and Differential Privacy | Even the most carefully curated economic data sets have variables that are noisy, missing, discretized, or privatized. The standard workflow for empirical research involves data cleaning followed by data analysis that typically ignores the bias and variance consequences of data cleaning. We formulate a semiparametric model for causal inference with corrupted data to encompass both data cleaning and data analysis. We propose a new end-to-end procedure for data cleaning, estimation, and inference with data cleaning-adjusted confidence intervals. We prove consistency, Gaussian approximation, and semiparametric efficiency for our estimator of the causal parameter by finite sample arguments. The rate of Gaussian approximation is $n^{-1/2}$ for global parameters such as average treatment effect, and it degrades gracefully for local parameters such as heterogeneous treatment effect for a specific demographic. Our key assumption is that the true covariates are approximately low rank. In our analysis, we provide nonasymptotic theoretical contributions to matrix completion, statistical learning, and semiparametric statistics. We verify the coverage of the data cleaning-adjusted confidence intervals in simulations calibrated to resemble differential privacy as implemented in the 2020 US Census. |
Automatic Estimation of Anthropometric Human Body Measurements | Research tasks related to human body analysis have been drawing a lot of attention in computer vision area over the last few decades, considering its potential benefits on our day-to-day life. Anthropometry is a field defining physical measures of a human body size, form, and functional capacities. Specifically, the accurate estimation of anthropometric body measurements from visual human body data is one of the challenging problems, where the solution would ease many different areas of applications, including ergonomics, garment manufacturing, etc. This paper formulates a research in the field of deep learning and neural networks, to tackle the challenge of body measurements estimation from various types of visual input data (such as 2D images or 3D point clouds). Also, we deal with the lack of real human data annotated with ground truth body measurements required for training and evaluation, by generating a synthetic dataset of various human body shapes and performing a skeleton-driven annotation. |
A Hybrid Algorithm for Convex Semidefinite Optimization | We present a hybrid algorithm for optimizing a convex, smooth function over the cone of positive semidefinite matrices. Our algorithm converges to the global optimal solution and can be used to solve general large-scale semidefinite programs and hence can be readily applied to a variety of machine learning problems. We show experimental results on three machine learning problems (matrix completion, metric learning, and sparse PCA) . Our approach outperforms state-of-the-art algorithms. |
Capacity, Bandwidth, and Compositionality in Emergent Language Learning | Many recent works have discussed the propensity, or lack thereof, for emergent languages to exhibit properties of natural languages. A favorite in the literature is learning compositionality. We note that most of those works have focused on communicative bandwidth as being of primary importance. While important, it is not the only contributing factor. In this paper, we investigate the learning biases that affect the efficacy and compositionality of emergent languages. Our foremost contribution is to explore how capacity of a neural network impacts its ability to learn a compositional language. We additionally introduce a set of evaluation metrics with which we analyze the learned languages. Our hypothesis is that there should be a specific range of model capacity and channel bandwidth that induces compositional structure in the resulting language and consequently encourages systematic generalization. While we empirically see evidence for the bottom of this range, we curiously do not find evidence for the top part of the range and believe that this is an open question for the community. |
Dynamic $\beta$-VAEs for quantifying biodiversity by clustering optically recorded insect signals | While insects are the largest and most diverse group of terrestrial animals, constituting ca. 80% of all known species, they are difficult to study due to their small size and similarity between species. Conventional monitoring techniques depend on time consuming trapping methods and tedious microscope-based work by skilled experts in order to identify the caught insect specimen at species, or even family level. Researchers and policy makers are in urgent need of a scalable monitoring tool in order to conserve biodiversity and secure human food production due to the rapid decline in insect numbers. In order to improve upon existing insect clustering methods, we propose an adaptive variant of the variational autoencoder (VAE) which is capable of clustering data by phylogenetic groups. The proposed dynamic beta-VAE dynamically adapts the scaling of the reconstruction and regularization loss terms (beta value) yielding useful latent representations of the input data. We demonstrate the usefulness of the dynamic beta-VAE on optically recorded insect signals from regions of southern Scandinavia to cluster unlabelled targets into possible species. We also demonstrate improved clustering performance in a semi-supervised setting using a small subset of labelled data. These experimental results, in both unsupervised- and semi-supervised settings, with the dynamic beta-VAE are promising and, in the near future, can be deployed to monitor insects and conserve the rapidly declining insect biodiversity. |
Gauge Equivariant Convolutional Networks and the Icosahedral CNN | The principle of equivariance to symmetry transformations enables a theoretically grounded approach to neural network architecture design. Equivariant networks have shown excellent performance and data efficiency on vision and medical imaging problems that exhibit symmetries. Here we show how this principle can be extended beyond global symmetries to local gauge transformations. This enables the development of a very general class of convolutional neural networks on manifolds that depend only on the intrinsic geometry, and which includes many popular methods from equivariant and geometric deep learning. We implement gauge equivariant CNNs for signals defined on the surface of the icosahedron, which provides a reasonable approximation of the sphere. By choosing to work with this very regular manifold, we are able to implement the gauge equivariant convolution using a single conv2d call, making it a highly scalable and practical alternative to Spherical CNNs. Using this method, we demonstrate substantial improvements over previous methods on the task of segmenting omnidirectional images and global climate patterns. |
Attention based Memory video portrait matting | We proposed a novel trimap free video matting method based on the attention mechanism. By the nature of the problem, most existing approaches use either multiple computational expansive modules or complex algorithms to exploit temporal information fully. We designed a temporal aggregation module to compute the temporal coherence between the current frame and its two previous frames. |
Identification of Bugs and Vulnerabilities in TLS Implementation for Windows Operating System Using State Machine Learning | TLS protocol is an essential part of secure Internet communication. In past, many attacks have been identified on the protocol. Most of these attacks are due to flaws in protocol implementation. The flaws are due to improper design and implementation of program logic by programmers. One of the widely used implementation of TLS is SChannel which is used in Windows operating system since its inception. We have used protocol state fuzzing to identify vulnerable and undesired state transitions in the state machine of the protocol for various versions of SChannel. The client as well as server components have been analyzed thoroughly using this technique and various flaws have been discovered in the implementation. Exploitation of these flaws under specific circumstances may lead to serious attacks which could disrupt secure communication. In this paper, we analyze state machine models of TLS protocol implementation of SChannel library and describe weaknesses and design flaws in these models, found using protocol state fuzzing. |
Online Enhanced Semantic Hashing: Towards Effective and Efficient Retrieval for Streaming Multi-Modal Data | With the vigorous development of multimedia equipment and applications, efficient retrieval of large-scale multi-modal data has become a trendy research topic. Thereinto, hashing has become a prevalent choice due to its retrieval efficiency and low storage cost. Although multi-modal hashing has drawn lots of attention in recent years, there still remain some problems. The first point is that existing methods are mainly designed in batch mode and not able to efficiently handle streaming multi-modal data. The second point is that all existing online multi-modal hashing methods fail to effectively handle unseen new classes which come continuously with streaming data chunks. In this paper, we propose a new model, termed Online enhAnced SemantIc haShing (OASIS). We design novel semantic-enhanced representation for data, which could help handle the new coming classes, and thereby construct the enhanced semantic objective function. An efficient and effective discrete online optimization algorithm is further proposed for OASIS. Extensive experiments show that our method can exceed the state-of-the-art models. For good reproducibility and benefiting the community, our code and data are already available in supplementary material and will be made publicly available. |
Software/Hardware Co-design for Multi-modal Multi-task Learning in Autonomous Systems | Optimizing the quality of result (QoR) and the quality of service (QoS) of AI-empowered autonomous systems simultaneously is very challenging. First, there are multiple input sources, e.g., multi-modal data from different sensors, requiring diverse data preprocessing, sensor fusion, and feature aggregation. Second, there are multiple tasks that require various AI models to run simultaneously, e.g., perception, localization, and control. Third, the computing and control system is heterogeneous, composed of hardware components with varied features, such as embedded CPUs, GPUs, FPGAs, and dedicated accelerators. Therefore, autonomous systems essentially require multi-modal multi-task (MMMT) learning which must be aware of hardware performance and implementation strategies. While MMMT learning has been attracting intensive research interests, its applications in autonomous systems are still underexplored. In this paper, we first discuss the opportunities of applying MMMT techniques in autonomous systems and then discuss the unique challenges that must be solved. In addition, we discuss the necessity and opportunities of MMMT model and hardware co-design, which is critical for autonomous systems especially with power/resource-limited or heterogeneous platforms. We formulate the MMMT model and heterogeneous hardware implementation co-design as a differentiable optimization problem, with the objective of improving the solution quality and reducing the overall power consumption and critical path latency. We advocate for further explorations of MMMT in autonomous systems and software/hardware co-design solutions. |
Prediction of Porosity and Permeability Alteration based on Machine Learning Algorithms | The objective of this work is to study the applicability of various Machine Learning algorithms for prediction of some rock properties which geoscientists usually define due to special lab analysis. We demonstrate that these special properties can be predicted only basing on routine core analysis (RCA) data. To validate the approach core samples from the reservoir with soluble rock matrix components (salts) were tested within 100+ laboratory experiments. The challenge of the experiments was to characterize the rate of salts in cores and alteration of porosity and permeability after reservoir desalination due to drilling mud or water injection. For these three measured characteristics, we developed the relevant predictive models, which were based on the results of RCA and data on coring depth and top and bottom depths of productive horizons. To select the most accurate Machine Learning algorithm a comparative analysis has been performed. It was shown that different algorithms work better in different models. However, two hidden layers Neural network has demonstrated the best predictive ability and generalizability for all three rock characteristics jointly. The other algorithms, such as Support Vector Machine and Linear Regression, also worked well on the dataset, but in particular cases. Overall, the applied approach allows predicting the alteration of porosity and permeability during desalination in porous rocks and also evaluating salt concentration without direct measurements in a laboratory. This work also shows that developed approaches could be applied for prediction of other rock properties (residual brine and oil saturations, relative permeability, capillary pressure, and others), which laboratory measurements are time-consuming and expensive. |
Modulation Pattern Detection Using Complex Convolutions in Deep Learning | Transceivers used for telecommunications transmit and receive specific modulation patterns that are represented as sequences of complex numbers. Classifying modulation patterns is challenging because noise and channel impairments affect the signals in complicated ways such that the received signal bears little resemblance to the transmitted signal. Although deep learning approaches have shown great promise over statistical methods in this problem space, deep learning frameworks continue to lag in support for complex-valued data. To address this gap, we study the implementation and use of complex convolutions in a series of convolutional neural network architectures. Replacement of data structure and convolution operations by their complex generalization in an architecture improves performance, with statistical significance, at recognizing modulation patterns in complex-valued signals with high SNR after being trained on low SNR signals. This suggests complex-valued convolutions enables networks to learn more meaningful representations. We investigate this hypothesis by comparing the features learned in each experiment by visualizing the inputs that results in one-hot modulation pattern classification for each network. |
F0 Modeling In Hmm-Based Speech Synthesis System Using Deep Belief Network | In recent years multilayer perceptrons (MLPs) with many hid- den layers Deep Neural Network (DNN) has performed sur- prisingly well in many speech tasks, i.e. speech recognition, speaker verification, speech synthesis etc. Although in the context of F0 modeling these techniques has not been ex- ploited properly. In this paper, Deep Belief Network (DBN), a class of DNN family has been employed and applied to model the F0 contour of synthesized speech which was generated by HMM-based speech synthesis system. The experiment was done on Bengali language. Several DBN-DNN architectures ranging from four to seven hidden layers and up to 200 hid- den units per hidden layer was presented and evaluated. The results were compared against clustering tree techniques pop- ularly found in statistical parametric speech synthesis. We show that from textual inputs DBN-DNN learns a high level structure which in turn improves F0 contour in terms of ob- jective and subjective tests. |
Learning to Pour | Pouring is a simple task people perform daily. It is the second most frequently executed motion in cooking scenarios, after pick-and-place. We present a pouring trajectory generation approach, which uses force feedback from the cup to determine the future velocity of pouring. The approach uses recurrent neural networks as its building blocks. We collected the pouring demonstrations which we used for training. To test our approach in simulation, we also created and trained a force estimation system. The simulated experiments show that the system is able to generalize to single unseen element of the pouring characteristics. |
Sample and Computationally Efficient Learning Algorithms under S-Concave Distributions | We provide new results for noise-tolerant and sample-efficient learning algorithms under $s$-concave distributions. The new class of $s$-concave distributions is a broad and natural generalization of log-concavity, and includes many important additional distributions, e.g., the Pareto distribution and $t$-distribution. This class has been studied in the context of efficient sampling, integration, and optimization, but much remains unknown about the geometry of this class of distributions and their applications in the context of learning. The challenge is that unlike the commonly used distributions in learning (uniform or more generally log-concave distributions), this broader class is not closed under the marginalization operator and many such distributions are fat-tailed. In this work, we introduce new convex geometry tools to study the properties of $s$-concave distributions and use these properties to provide bounds on quantities of interest to learning including the probability of disagreement between two halfspaces, disagreement outside a band, and the disagreement coefficient. We use these results to significantly generalize prior results for margin-based active learning, disagreement-based active learning, and passive learning of intersections of halfspaces. Our analysis of geometric properties of $s$-concave distributions might be of independent interest to optimization more broadly. |
Adversarial Examples for $k$-Nearest Neighbor Classifiers Based on Higher-Order Voronoi Diagrams | Adversarial examples are a widely studied phenomenon in machine learning models. While most of the attention has been focused on neural networks, other practical models also suffer from this issue. In this work, we propose an algorithm for evaluating the adversarial robustness of $k$-nearest neighbor classification, i.e., finding a minimum-norm adversarial example. Diverging from previous proposals, we take a geometric approach by performing a search that expands outwards from a given input point. On a high level, the search radius expands to the nearby Voronoi cells until we find a cell that classifies differently from the input point. To scale the algorithm to a large $k$, we introduce approximation steps that find perturbations with smaller norm, compared to the baselines, in a variety of datasets. Furthermore, we analyze the structural properties of a dataset where our approach outperforms the competition. |
Metadata-Based Detection of Child Sexual Abuse Material | Child Sexual Abuse Media (CSAM) is any visual record of a sexually-explicit activity involving minors. CSAM impacts victims differently from the actual abuse because the distribution never ends, and images are permanent. Machine learning-based solutions can help law enforcement quickly identify CSAM and block digital distribution. However, collecting CSAM imagery to train machine learning models has many ethical and legal constraints, creating a barrier to research development. With such restrictions in place, the development of CSAM machine learning detection systems based on file metadata uncovers several opportunities. Metadata is not a record of a crime, and it does not have legal restrictions. Therefore, investing in detection systems based on metadata can increase the rate of discovery of CSAM and help thousands of victims. We propose a framework for training and evaluating deployment-ready machine learning models for CSAM identification. Our framework provides guidelines to evaluate CSAM detection models against intelligent adversaries and models' performance with open data. We apply the proposed framework to the problem of CSAM detection based on file paths. In our experiments, the best-performing model is based on convolutional neural networks and achieves an accuracy of 0.97. Our evaluation shows that the CNN model is robust against offenders actively trying to evade detection by evaluating the model against adversarially modified data. Experiments with open datasets confirm that the model generalizes well and is deployment-ready. |
A Deep Learning Approach for Thermal Plume Prediction of Groundwater Heat Pumps | Climate control of buildings makes up a significant portion of global energy consumption, with groundwater heat pumps providing a suitable alternative. To prevent possibly negative interactions between heat pumps throughout a city, city planners have to optimize their layouts in the future. We develop a novel data-driven approach for building small-scale surrogates for modelling the thermal plumes generated by groundwater heat pumps in the surrounding subsurface water. Building on a data set generated from 2D numerical simulations, we train a convolutional neural network for predicting steady-state subsurface temperature fields from a given subsurface velocity field. We show that compared to existing models ours can capture more complex dynamics while still being quick to compute. The resulting surrogate is thus well-suited for interactive design tools by city planners. |
Reducing Information Bottleneck for Weakly Supervised Semantic Segmentation | Weakly supervised semantic segmentation produces pixel-level localization from class labels; however, a classifier trained on such labels is likely to focus on a small discriminative region of the target object. We interpret this phenomenon using the information bottleneck principle: the final layer of a deep neural network, activated by the sigmoid or softmax activation functions, causes an information bottleneck, and as a result, only a subset of the task-relevant information is passed on to the output. We first support this argument through a simulated toy experiment and then propose a method to reduce the information bottleneck by removing the last activation function. In addition, we introduce a new pooling method that further encourages the transmission of information from non-discriminative regions to the classification. Our experimental evaluations demonstrate that this simple modification significantly improves the quality of localization maps on both the PASCAL VOC 2012 and MS COCO 2014 datasets, exhibiting a new state-of-the-art performance for weakly supervised semantic segmentation. The code is available at: https://github.com/jbeomlee93/RIB. |
Decomposable-Net: Scalable Low-Rank Compression for Neural Networks | Compressing DNNs is important for the real-world applications operating on resource-constrained devices. However, we typically observe drastic performance deterioration when changing model size after training is completed. Therefore, retraining is required to resume the performance of the compressed models suitable for different devices. In this paper, we propose Decomposable-Net (the network decomposable in any size), which allows flexible changes to model size without retraining. We decompose weight matrices in the DNNs via singular value decomposition and adjust ranks according to the target model size. Unlike the existing low-rank compression methods that specialize the model to a fixed size, we propose a novel backpropagation scheme that jointly minimizes losses for both of full- and low-rank networks. This enables not only to maintain the performance of a full-rank network {\it without retraining} but also to improve low-rank networks in multiple sizes. Additionally, we introduce a simple criterion for rank selection that effectively suppresses approximation error. In experiments on the ImageNet classification task, Decomposable-Net yields superior accuracy in a wide range of model sizes. In particular, Decomposable-Net achieves the top-1 accuracy of $73.2\%$ with $0.27\times$MACs with ResNet-50, compared to Tucker decomposition ($67.4\% / 0.30\times$), Trained Rank Pruning ($70.6\% / 0.28\times$), and universally slimmable networks ($71.4\% / 0.26\times$). |
Log-DenseNet: How to Sparsify a DenseNet | Skip connections are increasingly utilized by deep neural networks to improve accuracy and cost-efficiency. In particular, the recent DenseNet is efficient in computation and parameters, and achieves state-of-the-art predictions by directly connecting each feature layer to all previous ones. However, DenseNet's extreme connectivity pattern may hinder its scalability to high depths, and in applications like fully convolutional networks, full DenseNet connections are prohibitively expensive. This work first experimentally shows that one key advantage of skip connections is to have short distances among feature layers during backpropagation. Specifically, using a fixed number of skip connections, the connection patterns with shorter backpropagation distance among layers have more accurate predictions. Following this insight, we propose a connection template, Log-DenseNet, which, in comparison to DenseNet, only slightly increases the backpropagation distances among layers from 1 to ($1 + \log_2 L$), but uses only $L\log_2 L$ total connections instead of $O(L^2)$. Hence, Log-DenseNets are easier than DenseNets to implement and to scale. We demonstrate the effectiveness of our design principle by showing better performance than DenseNets on tabula rasa semantic segmentation, and competitive results on visual recognition. |
Adaptive Subcarrier, Parameter, and Power Allocation for Partitioned Edge Learning Over Broadband Channels | In this paper, we consider partitioned edge learning (PARTEL), which implements parameter-server training, a well known distributed learning method, in a wireless network. Thereby, PARTEL leverages distributed computation resources at edge devices to train a large-scale artificial intelligence (AI) model by dynamically partitioning the model into parametric blocks for separated updating at devices. Targeting broadband channels, we consider the joint control of parameter allocation, sub-channel allocation, and transmission power to improve the performance of PARTEL. Specifically, the policies for joint SUbcarrier, Parameter, and POweR allocaTion (SUPPORT) are optimized under the criterion of minimum learning latency. Two cases are considered. First, for the case of decomposable models (e.g., logistic regression), the latency-minimization problem is a mixed-integer program and non-convex. Due to its intractability, we develop a practical solution by integer relaxation and transforming it into an equivalent convex problem of model size maximization under a latency constraint. Thereby, a low-complexity algorithm is designed to compute the SUPPORT policy. Second, consider the case of deep neural network (DNN) models which can be trained using PARTEL by introducing some auxiliary variables. This, however, introduces constraints on model partitioning reducing the granularity of parameter allocation. The preceding policy is extended to DNN models by applying the proposed techniques of load rounding and proportional adjustment to rein in latency expansion caused by the load granularity constraints. |
Incorporating Sum Constraints into Multitask Gaussian Processes | Machine learning models can be improved by adapting them to respect existing background knowledge. In this paper we consider multitask Gaussian processes, with background knowledge in the form of constraints that require a specific sum of the outputs to be constant. This is achieved by conditioning the prior distribution on the constraint fulfillment. The approach allows for both linear and nonlinear constraints. We demonstrate that the constraints are fulfilled with high precision and that the construction can improve the overall prediction accuracy as compared to the standard Gaussian process. |
Mutual Learning Network for Multi-Source Domain Adaptation | Early Unsupervised Domain Adaptation (UDA) methods have mostly assumed the setting of a single source domain, where all the labeled source data come from the same distribution. However, in practice the labeled data can come from multiple source domains with different distributions. In such scenarios, the single source domain adaptation methods can fail due to the existence of domain shifts across different source domains and multi-source domain adaptation methods need to be designed. In this paper, we propose a novel multi-source domain adaptation method, Mutual Learning Network for Multiple Source Domain Adaptation (ML-MSDA). Under the framework of mutual learning, the proposed method pairs the target domain with each single source domain to train a conditional adversarial domain adaptation network as a branch network, while taking the pair of the combined multi-source domain and target domain to train a conditional adversarial adaptive network as the guidance network. The multiple branch networks are aligned with the guidance network to achieve mutual learning by enforcing JS-divergence regularization over their prediction probability distributions on the corresponding target data. We conduct extensive experiments on multiple multi-source domain adaptation benchmark datasets. The results show the proposed ML-MSDA method outperforms the comparison methods and achieves the state-of-the-art performance. |
A SPA-based Manifold Learning Framework for Motor Imagery EEG Data Classification | The electroencephalography (EEG) signal is a non-stationary, stochastic, and highly non-linear bioelectric signal for which achieving high classification accuracy is challenging, especially when the number of subjects is limited. As frequently used solution, classifiers based on multilayer neural networks has to be implemented without large training data sets and careful tuning. This paper proposes a manifold learning framework to classify two types of EEG data from motor imagery (MI) tasks by discovering lower dimensional geometric structures. For feature extraction, it is implemented by Common Spatial Pattern (CSP) from the preprocessed EEG signals. In the neighborhoods of the features for classification, the local approximation to the support of the data is obtained, and then the features are assigned to the classes with the closest support. A spherical approximation (SPA) classifier is created using spherelets for local approximation, and the extracted features are classified with this manifold-based method. The SPA classifier achieves high accuracy in the 2008 BCI competition data, and the analysis shows that this method can significantly improve the decoding accuracy of MI tasks and exhibit strong robustness for small sample datasets. It would be simple and efficient to tune the two-parameters classifier for the online brain-computer interface(BCI)system. |
Decoupling Exploration and Exploitation for Meta-Reinforcement Learning without Sacrifices | The goal of meta-reinforcement learning (meta-RL) is to build agents that can quickly learn new tasks by leveraging prior experience on related tasks. Learning a new task often requires both exploring to gather task-relevant information and exploiting this information to solve the task. In principle, optimal exploration and exploitation can be learned end-to-end by simply maximizing task performance. However, such meta-RL approaches struggle with local optima due to a chicken-and-egg problem: learning to explore requires good exploitation to gauge the exploration's utility, but learning to exploit requires information gathered via exploration. Optimizing separate objectives for exploration and exploitation can avoid this problem, but prior meta-RL exploration objectives yield suboptimal policies that gather information irrelevant to the task. We alleviate both concerns by constructing an exploitation objective that automatically identifies task-relevant information and an exploration objective to recover only this information. This avoids local optima in end-to-end training, without sacrificing optimal exploration. Empirically, DREAM substantially outperforms existing approaches on complex meta-RL problems, such as sparse-reward 3D visual navigation. Videos of DREAM: https://ezliu.github.io/dream/ |
MATCH-Net: Dynamic Prediction in Survival Analysis using Convolutional Neural Networks | Accurate prediction of disease trajectories is critical for early identification and timely treatment of patients at risk. Conventional methods in survival analysis are often constrained by strong parametric assumptions and limited in their ability to learn from high-dimensional data, while existing neural network models are not readily-adapted to the longitudinal setting. This paper develops a novel convolutional approach that addresses these drawbacks. We present MATCH-Net: a Missingness-Aware Temporal Convolutional Hitting-time Network, designed to capture temporal dependencies and heterogeneous interactions in covariate trajectories and patterns of missingness. To the best of our knowledge, this is the first investigation of temporal convolutions in the context of dynamic prediction for personalized risk prognosis. Using real-world data from the Alzheimer's Disease Neuroimaging Initiative, we demonstrate state-of-the-art performance without making any assumptions regarding underlying longitudinal or time-to-event processes attesting to the model's potential utility in clinical decision support. |
A Two-Stage Federated Transfer Learning Framework in Medical Images Classification on Limited Data: A COVID-19 Case Study | COVID-19 pandemic has spread rapidly and caused a shortage of global medical resources. The efficiency of COVID-19 diagnosis has become highly significant. As deep learning and convolutional neural network (CNN) has been widely utilized and been verified in analyzing medical images, it has become a powerful tool for computer-assisted diagnosis. However, there are two most significant challenges in medical image classification with the help of deep learning and neural networks, one of them is the difficulty of acquiring enough samples, which may lead to model overfitting. Privacy concerns mainly bring the other challenge since medical-related records are often deemed patients' private information and protected by laws such as GDPR and HIPPA. Federated learning can ensure the model training is decentralized on different devices and no data is shared among them, which guarantees privacy. However, with data located on different devices, the accessible data of each device could be limited. Since transfer learning has been verified in dealing with limited data with good performance, therefore, in this paper, We made a trial to implement federated learning and transfer learning techniques using CNNs to classify COVID-19 using lung CT scans. We also explored the impact of dataset distribution at the client-side in federated learning and the number of training epochs a model is trained. Finally, we obtained very high performance with federated learning, demonstrating our success in leveraging accuracy and privacy. |
Haar Graph Pooling | Deep Graph Neural Networks (GNNs) are useful models for graph classification and graph-based regression tasks. In these tasks, graph pooling is a critical ingredient by which GNNs adapt to input graphs of varying size and structure. We propose a new graph pooling operation based on compressive Haar transforms -- HaarPooling. HaarPooling implements a cascade of pooling operations; it is computed by following a sequence of clusterings of the input graph. A HaarPooling layer transforms a given input graph to an output graph with a smaller node number and the same feature dimension; the compressive Haar transform filters out fine detail information in the Haar wavelet domain. In this way, all the HaarPooling layers together synthesize the features of any given input graph into a feature vector of uniform size. Such transforms provide a sparse characterization of the data and preserve the structure information of the input graph. GNNs implemented with standard graph convolution layers and HaarPooling layers achieve state of the art performance on diverse graph classification and regression problems. |
Hidden Markov Models with Momentum | Momentum is a popular technique for improving convergence rates during gradient descent. In this research, we experiment with adding momentum to the Baum-Welch expectation-maximization algorithm for training Hidden Markov Models. We compare discrete Hidden Markov Models trained with and without momentum on English text and malware opcode data. The effectiveness of momentum is determined by measuring the changes in model score and classification accuracy due to momentum. Our extensive experiments indicate that adding momentum to Baum-Welch can reduce the number of iterations required for initial convergence during HMM training, particularly in cases where the model is slow to converge. However, momentum does not seem to improve the final model performance at a high number of iterations. |
Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN) | A central question of representation learning asks under which conditions it is possible to reconstruct the true latent variables of an arbitrarily complex generative process. Recent breakthrough work by Khemakhem et al. (2019) on nonlinear ICA has answered this question for a broad class of conditional generative processes. We extend this important result in a direction relevant for application to real-world data. First, we generalize the theory to the case of unknown intrinsic problem dimension and prove that in some special (but not very restrictive) cases, informative latent variables will be automatically separated from noise by an estimating model. Furthermore, the recovered informative latent variables will be in one-to-one correspondence with the true latent variables of the generating process, up to a trivial component-wise transformation. Second, we introduce a modification of the RealNVP invertible neural network architecture (Dinh et al. (2016)) which is particularly suitable for this type of problem: the General Incompressible-flow Network (GIN). Experiments on artificial data and EMNIST demonstrate that theoretical predictions are indeed verified in practice. In particular, we provide a detailed set of exactly 22 informative latent variables extracted from EMNIST. |
COVID-19 Cough Classification using Machine Learning and Global Smartphone Recordings | We present a machine learning based COVID-19 cough classifier which can discriminate COVID-19 positive coughs from both COVID-19 negative and healthy coughs recorded on a smartphone. This type of screening is non-contact, easy to apply, and can reduce the workload in testing centres as well as limit transmission by recommending early self-isolation to those who have a cough suggestive of COVID-19. The datasets used in this study include subjects from all six continents and contain both forced and natural coughs, indicating that the approach is widely applicable. The publicly available Coswara dataset contains 92 COVID-19 positive and 1079 healthy subjects, while the second smaller dataset was collected mostly in South Africa and contains 18 COVID-19 positive and 26 COVID-19 negative subjects who have undergone a SARS-CoV laboratory test. Both datasets indicate that COVID-19 positive coughs are 15\%-20\% shorter than non-COVID coughs. Dataset skew was addressed by applying the synthetic minority oversampling technique (SMOTE). A leave-$p$-out cross-validation scheme was used to train and evaluate seven machine learning classifiers: LR, KNN, SVM, MLP, CNN, LSTM and Resnet50. Our results show that although all classifiers were able to identify COVID-19 coughs, the best performance was exhibited by the Resnet50 classifier, which was best able to discriminate between the COVID-19 positive and the healthy coughs with an area under the ROC curve (AUC) of 0.98. An LSTM classifier was best able to discriminate between the COVID-19 positive and COVID-19 negative coughs, with an AUC of 0.94 after selecting the best 13 features from a sequential forward selection (SFS). Since this type of cough audio classification is cost-effective and easy to deploy, it is potentially a useful and viable means of non-contact COVID-19 screening. |
Attributional Robustness Training using Input-Gradient Spatial Alignment | Interpretability is an emerging area of research in trustworthy machine learning. Safe deployment of machine learning system mandates that the prediction and its explanation be reliable and robust. Recently, it has been shown that the explanations could be manipulated easily by adding visually imperceptible perturbations to the input while keeping the model's prediction intact. In this work, we study the problem of attributional robustness (i.e. models having robust explanations) by showing an upper bound for attributional vulnerability in terms of spatial correlation between the input image and its explanation map. We propose a training methodology that learns robust features by minimizing this upper bound using soft-margin triplet loss. Our methodology of robust attribution training (\textit{ART}) achieves the new state-of-the-art attributional robustness measure by a margin of $\approx$ 6-18 $\%$ on several standard datasets, ie. SVHN, CIFAR-10 and GTSRB. We further show the utility of the proposed robust training technique (\textit{ART}) in the downstream task of weakly supervised object localization by achieving the new state-of-the-art performance on CUB-200 dataset. |
Stochastic Deconvolutional Neural Network Ensemble Training on Generative Pseudo-Adversarial Networks | The training of Generative Adversarial Networks is a difficult task mainly due to the nature of the networks. One such issue is when the generator and discriminator start oscillating, rather than converging to a fixed point. Another case can be when one agent becomes more adept than the other which results in the decrease of the other agent's ability to learn, reducing the learning capacity of the system as a whole. Additionally, there exists the problem of Mode Collapse which involves the generators output collapsing to a single sample or a small set of similar samples. To train GANs a careful selection of the architecture that is used along with a variety of other methods to improve training. Even when applying these methods there is low stability of training in relation to the parameters that are chosen. Stochastic ensembling is suggested as a method for improving the stability while training GANs. |
Physics-aware deep neural networks for surrogate modeling of turbulent natural convection | Recent works have explored the potential of machine learning as data-driven turbulence closures for RANS and LES techniques. Beyond these advances, the high expressivity and agility of physics-informed neural networks (PINNs) make them promising candidates for full fluid flow PDE modeling. An important question is whether this new paradigm, exempt from the traditional notion of discretization of the underlying operators very much connected to the flow scales resolution, is capable of sustaining high levels of turbulence characterized by multi-scale features? We investigate the use of PINNs surrogate modeling for turbulent Rayleigh-B{\'e}nard (RB) convection flows in rough and smooth rectangular cavities, mainly relying on DNS temperature data from the fluid bulk. We carefully quantify the computational requirements under which the formulation is capable of accurately recovering the flow hidden quantities. We then propose a new padding technique to distribute some of the scattered coordinates-at which PDE residuals are minimized-around the region of labeled data acquisition. We show how it comes to play as a regularization close to the training boundaries which are zones of poor accuracy for standard PINNs and results in a noticeable global accuracy improvement at iso-budget. Finally, we propose for the first time to relax the incompressibility condition in such a way that it drastically benefits the optimization search and results in a much improved convergence of the composite loss function. The RB results obtained at high Rayleigh number Ra = 2 $\bullet$ 10 9 are particularly impressive: the predictive accuracy of the surrogate over the entire half a billion DNS coordinates yields errors for all flow variables ranging between [0.3% -- 4%] in the relative L 2 norm, with a training relying only on 1.6% of the DNS data points. |
Radar Human Motion Recognition Using Motion States and Two-Way Classifications | We perform classification of activities of daily living (ADL) using a Frequency-Modulated Continuous Waveform (FMCW) radar. In particular, we consider contiguous motions that are inseparable in time. Both the micro-Doppler signature and range-map are used to determine transitions from translation (walking) to in-place motions and vice versa, as well as to provide motion onset and the offset times. The possible classes of activities post and prior to the translation motion can be separately handled by forward and background classifiers. The paper describes ADL in terms of states and transitioning actions, and sets a framework to deal with separable and inseparable contiguous motions. It is shown that considering only the physically possible classes of motions stemming from the current motion state improves classification rates compared to incorporating all ADL for any given time. |
Tensor Decompositions for Hyperspectral Data Processing in Remote Sensing: A Comprehensive Review | Owing to the rapid development of sensor technology, hyperspectral (HS) remote sensing (RS) imaging has provided a significant amount of spatial and spectral information for the observation and analysis of the Earth's surface at a distance of data acquisition devices, such as aircraft, spacecraft, and satellite. The recent advancement and even revolution of the HS RS technique offer opportunities to realize the full potential of various applications, while confronting new challenges for efficiently processing and analyzing the enormous HS acquisition data. Due to the maintenance of the 3-D HS inherent structure, tensor decomposition has aroused widespread concern and research in HS data processing tasks over the past decades. In this article, we aim at presenting a comprehensive overview of tensor decomposition, specifically contextualizing the five broad topics in HS data processing, and they are HS restoration, compressed sensing, anomaly detection, super-resolution, and spectral unmixing. For each topic, we elaborate on the remarkable achievements of tensor decomposition models for HS RS with a pivotal description of the existing methodologies and a representative exhibition on the experimental results. As a result, the remaining challenges of the follow-up research directions are outlined and discussed from the perspective of the real HS RS practices and tensor decomposition merged with advanced priors and even with deep neural networks. This article summarizes different tensor decomposition-based HS data processing methods and categorizes them into different classes from simple adoptions to complex combinations with other priors for the algorithm beginners. We also expect this survey can provide new investigations and development trends for the experienced researchers who understand tensor decomposition and HS RS to some extent. |
ConvDySAT: Deep Neural Representation Learning on Dynamic Graphs via Self-Attention and Convolutional Neural Networks | Learning node representations on temporal graphs is a fundamental step to learn real-word dynamic graphs efficiently. Real-world graphs have the nature of continuously evolving over time, such as changing edges weights, removing and adding nodes and appearing and disappearing of edges, while previous graph representation learning methods focused generally on static graphs. We present ConvDySAT as an enhancement of DySAT, one of the state-of-the-art dynamic methods, by augmenting convolution neural networks with the self-attention mechanism, the employed method in DySAT to express the structural and temporal evolution. We conducted single-step link prediction on a communication network and rating network, Experimental results show significant performance gains for ConvDySAT over various state-of-the-art methods. |
PR-DAD: Phase Retrieval Using Deep Auto-Decoders | Phase retrieval is a well known ill-posed inverse problem where one tries to recover images given only the magnitude values of their Fourier transform as input. In recent years, new algorithms based on deep learning have been proposed, providing breakthrough results that surpass the results of the classical methods. In this work we provide a novel deep learning architecture PR-DAD (Phase Retrieval Using Deep Auto- Decoders), whose components are carefully designed based on mathematical modeling of the phase retrieval problem. The architecture provides experimental results that surpass all current results. |
Exact Shapley Values for Local and Model-True Explanations of Decision Tree Ensembles | Additive feature explanations using Shapley values have become popular for providing transparency into the relative importance of each feature to an individual prediction of a machine learning model. While Shapley values provide a unique additive feature attribution in cooperative game theory, the Shapley values that can be generated for even a single machine learning model are far from unique, with theoretical and implementational decisions affecting the resulting attributions. Here, we consider the application of Shapley values for explaining decision tree ensembles and present a novel approach to Shapley value-based feature attribution that can be applied to random forests and boosted decision trees. This new method provides attributions that accurately reflect details of the model prediction algorithm for individual instances, while being computationally competitive with one of the most widely used current methods. We explain the theoretical differences between the standard and novel approaches and compare their performance using synthetic and real data. |
BooST: Boosting Smooth Trees for Partial Effect Estimation in Nonlinear Regressions | In this paper, we introduce a new machine learning (ML) model for nonlinear regression called the Boosted Smooth Transition Regression Trees (BooST), which is a combination of boosting algorithms with smooth transition regression trees. The main advantage of the BooST model is the estimation of the derivatives (partial effects) of very general nonlinear models. Therefore, the model can provide more interpretation about the mapping between the covariates and the dependent variable than other tree-based models, such as Random Forests. We present several examples with both simulated and real data. |
Understanding Overparameterization in Generative Adversarial Networks | A broad class of unsupervised deep learning methods such as Generative Adversarial Networks (GANs) involve training of overparameterized models where the number of parameters of the model exceeds a certain threshold. A large body of work in supervised learning have shown the importance of model overparameterization in the convergence of the gradient descent (GD) to globally optimal solutions. In contrast, the unsupervised setting and GANs in particular involve non-convex concave mini-max optimization problems that are often trained using Gradient Descent/Ascent (GDA). The role and benefits of model overparameterization in the convergence of GDA to a global saddle point in non-convex concave problems is far less understood. In this work, we present a comprehensive analysis of the importance of model overparameterization in GANs both theoretically and empirically. We theoretically show that in an overparameterized GAN model with a $1$-layer neural network generator and a linear discriminator, GDA converges to a global saddle point of the underlying non-convex concave min-max problem. To the best of our knowledge, this is the first result for global convergence of GDA in such settings. Our theory is based on a more general result that holds for a broader class of nonlinear generators and discriminators that obey certain assumptions (including deeper generators and random feature discriminators). We also empirically study the role of model overparameterization in GANs using several large-scale experiments on CIFAR-10 and Celeb-A datasets. Our experiments show that overparameterization improves the quality of generated samples across various model architectures and datasets. Remarkably, we observe that overparameterization leads to faster and more stable convergence behavior of GDA across the board. |
Bayesian experimental design using regularized determinantal point processes | In experimental design, we are given $n$ vectors in $d$ dimensions, and our goal is to select $k\ll n$ of them to perform expensive measurements, e.g., to obtain labels/responses, for a linear regression task. Many statistical criteria have been proposed for choosing the optimal design, with popular choices including A- and D-optimality. If prior knowledge is given, typically in the form of a $d\times d$ precision matrix $\mathbf A$, then all of the criteria can be extended to incorporate that information via a Bayesian framework. In this paper, we demonstrate a new fundamental connection between Bayesian experimental design and determinantal point processes, the latter being widely used for sampling diverse subsets of data. We use this connection to develop new efficient algorithms for finding $(1+\epsilon)$-approximations of optimal designs under four optimality criteria: A, C, D and V. Our algorithms can achieve this when the desired subset size $k$ is $\Omega(\frac{d_{\mathbf A}}{\epsilon} + \frac{\log 1/\epsilon}{\epsilon^2})$, where $d_{\mathbf A}\leq d$ is the $\mathbf A$-effective dimension, which can often be much smaller than $d$. Our results offer direct improvements over a number of prior works, for both Bayesian and classical experimental design, in terms of algorithm efficiency, approximation quality, and range of applicable criteria. |
Cultural association based on machine learning for team formation | Culture is core to human civilization, and is essential for human intellectual achievements in social context. Culture also influences how humans work together, perform particular task and overall lifestyle and dealing with other groups of civilization. Thus, culture is concerned with establishing shared ideas, particularly those playing a key role in success. Does it impact on how two individuals can work together in achieving certain goals? In this paper, we establish a means to derive cultural association and map it to culturally mediated success. Human interactions with the environment are typically in the form of expressions. Association between culture and behavior produce similar beliefs which lead to common principles and actions, while cultural similarity as a set of common expressions and responses. To measure cultural association among different candidates, we propose the use of a Graphical Association Method (GAM). The behaviors of candidates are captured through series of expressions and represented in the graphical form. The association among corresponding node and core nodes is used for the same. Our approach provides a number of interesting results and promising avenues for future applications. |
Scalable Statistical Inference of Photometric Redshift via Data Subsampling | Handling big data has largely been a major bottleneck in traditional statistical models. Consequently, when accurate point prediction is the primary target, machine learning models are often preferred over their statistical counterparts for bigger problems. But full probabilistic statistical models often outperform other models in quantifying uncertainties associated with model predictions. We develop a data-driven statistical modeling framework that combines the uncertainties from an ensemble of statistical models learned on smaller subsets of data carefully chosen to account for imbalances in the input space. We demonstrate this method on a photometric redshift estimation problem in cosmology, which seeks to infer a distribution of the redshift -- the stretching effect in observing the light of far-away galaxies -- given multivariate color information observed for an object in the sky. Our proposed method performs balanced partitioning, graph-based data subsampling across the partitions, and training of an ensemble of Gaussian process models. |
Byzantine-Robust Decentralized Stochastic Optimization over Static and Time-Varying Networks | In this paper, we consider the Byzantine-robust stochastic optimization problem defined over decentralized static and time-varying networks, where the agents collaboratively minimize the summation of expectations of stochastic local cost functions, but some of the agents are unreliable due to data corruptions, equipment failures or cyber-attacks. The unreliable agents, which are called as Byzantine agents thereafter, can send faulty values to their neighbors and bias the optimization process. Our key idea to handle the Byzantine attacks is to formulate a total variation (TV) norm-penalized approximation of the Byzantine-free problem, where the penalty term forces the local models of regular agents to be close, but also allows the existence of outliers from the Byzantine agents. A stochastic subgradient method is applied to solve the penalized problem. We prove that the proposed method reaches a neighborhood of the Byzantine-free optimal solution, and the size of neighborhood is determined by the number of Byzantine agents and the network topology. Numerical experiments corroborate the theoretical analysis, as well as demonstrate the robustness of the proposed method to Byzantine attacks and its superior performance comparing to existing methods. |
How and Why to Use Experimental Data to Evaluate Methods for Observational Causal Inference | Methods that infer causal dependence from observational data are central to many areas of science, including medicine, economics, and the social sciences. A variety of theoretical properties of these methods have been proven, but empirical evaluation remains a challenge, largely due to the lack of observational data sets for which treatment effect is known. We describe and analyze observational sampling from randomized controlled trials (OSRCT), a method for evaluating causal inference methods using data from randomized controlled trials (RCTs). This method can be used to create constructed observational data sets with corresponding unbiased estimates of treatment effect, substantially increasing the number of data sets available for empirical evaluation of causal inference methods. We show that, in expectation, OSRCT creates data sets that are equivalent to those produced by randomly sampling from empirical data sets in which all potential outcomes are available. We then perform a large-scale evaluation of seven causal inference methods over 37 data sets, drawn from RCTs, as well as simulators, real-world computational systems, and observational data sets augmented with a synthetic response variable. We find notable performance differences when comparing across data from different sources, demonstrating the importance of using data from a variety of sources when evaluating any causal inference method. |
Batch Tournament Selection for Genetic Programming | Lexicase selection achieves very good solution quality by introducing ordered test cases. However, the computational complexity of lexicase selection can prohibit its use in many applications. In this paper, we introduce Batch Tournament Selection (BTS), a hybrid of tournament and lexicase selection which is approximately one order of magnitude faster than lexicase selection while achieving a competitive quality of solutions. Tests on a number of regression datasets show that BTS compares well with lexicase selection in terms of mean absolute error while having a speed-up of up to 25 times. Surprisingly, BTS and lexicase selection have almost no difference in both diversity and performance. This reveals that batches and ordered test cases are completely different mechanisms which share the same general principle fostering the specialization of individuals. This work introduces an efficient algorithm that sheds light onto the main principles behind the success of lexicase, potentially opening up a new range of possibilities for algorithms to come. |
Functional Ensemble Distillation | Bayesian models have many desirable properties, most notable is their ability to generalize from limited data and to properly estimate the uncertainty in their predictions. However, these benefits come at a steep computational cost as Bayesian inference, in most cases, is computationally intractable. One popular approach to alleviate this problem is using a Monte-Carlo estimation with an ensemble of models sampled from the posterior. However, this approach still comes at a significant computational cost, as one needs to store and run multiple models at test time. In this work, we investigate how to best distill an ensemble's predictions using an efficient model. First, we argue that current approaches that simply return distribution over predictions cannot compute important properties, such as the covariance between predictions, which can be valuable for further processing. Second, in many limited data settings, all ensemble members achieve nearly zero training loss, namely, they produce near-identical predictions on the training set which results in sub-optimal distilled models. To address both problems, we propose a novel and general distillation approach, named Functional Ensemble Distillation (FED), and we investigate how to best distill an ensemble in this setting. We find that learning the distilled model via a simple augmentation scheme in the form of mixup augmentation significantly boosts the performance. We evaluated our method on several tasks and showed that it achieves superior results in both accuracy and uncertainty estimation compared to current approaches. |
AdamP: Slowing Down the Slowdown for Momentum Optimizers on Scale-invariant Weights | Normalization techniques are a boon for modern deep learning. They let weights converge more quickly with often better generalization performances. It has been argued that the normalization-induced scale invariance among the weights provides an advantageous ground for gradient descent (GD) optimizers: the effective step sizes are automatically reduced over time, stabilizing the overall training procedure. It is often overlooked, however, that the additional introduction of momentum in GD optimizers results in a far more rapid reduction in effective step sizes for scale-invariant weights, a phenomenon that has not yet been studied and may have caused unwanted side effects in the current practice. This is a crucial issue because arguably the vast majority of modern deep neural networks consist of (1) momentum-based GD (e.g. SGD or Adam) and (2) scale-invariant parameters. In this paper, we verify that the widely-adopted combination of the two ingredients lead to the premature decay of effective step sizes and sub-optimal model performances. We propose a simple and effective remedy, SGDP and AdamP: get rid of the radial component, or the norm-increasing direction, at each optimizer step. Because of the scale invariance, this modification only alters the effective step sizes without changing the effective update directions, thus enjoying the original convergence properties of GD optimizers. Given the ubiquity of momentum GD and scale invariance in machine learning, we have evaluated our methods against the baselines on 13 benchmarks. They range from vision tasks like classification (e.g. ImageNet), retrieval (e.g. CUB and SOP), and detection (e.g. COCO) to language modelling (e.g. WikiText) and audio classification (e.g. DCASE) tasks. We verify that our solution brings about uniform gains in those benchmarks. Source code is available at https://github.com/clovaai/AdamP. |
Revealing the real-world CO2 emission reduction of ridesplitting and its determinants based on machine learning | Ridesplitting, which is a form of pooled ridesourcing service, has great potential to alleviate the negative impacts of ridesourcing on the environment. However, most existing studies only explored its theoretical environmental benefits based on optimization models and simulations. To put into practice, this study aims to reveal the real-world emission reduction of ridesplitting and its determinants based on the observed data of ridesourcing in Chengdu, China. Integrating the trip data with the COPERT model, this study calculates the CO2 emissions of shared rides (ridesplitting) and their substituted single rides (regular ridesourcing) to estimate the CO2 emission reduction of each ridesplitting trip. The results show that not all ridesplitting trips reduce emissions from ridesourcing in the real world. The CO2 emission reduction rate of ridesplitting varies from trip to trip, averaging at 43.15g/km. Then, the interpretable machine learning models, gradient boosting machines, are applied to explore the relationship between the CO2 emission reduction rate of ridesplitting and its determinants. Based on the SHapley Additive exPlanations method, the overlap rate and detour rate of shared rides are identified to be the most important factors that determine the CO2 emission reduction rate of ridesplitting. Increasing the overlap rate, the number of shared rides, average speed, and ride distance ratio and decreasing the detour rate, actual trip distance, ride distance gap can increase the CO2 emission reduction rate of ridesplitting. In addition, nonlinear effects and interactions of several key factors are examined through the partial dependence plots. This study provides a scientific method for the government and ridesourcing companies to better assess and optimize the environmental benefits of ridesplitting. |
Improving Mechanical Ventilator Clinical Decision Support Systems with A Machine Learning Classifier for Determining Ventilator Mode | Clinical decision support systems (CDSS) will play an in-creasing role in improving the quality of medical care for critically ill patients. However, due to limitations in current informatics infrastructure, CDSS do not always have com-plete information on state of supporting physiologic monitor-ing devices, which can limit the input data available to CDSS. This is especially true in the use case of mechanical ventilation (MV), where current CDSS have no knowledge of critical ventilation settings, such as ventilation mode. To enable MV CDSS to make accurate recommendations related to ventilator mode, we developed a highly performant ma-chine learning model that is able to perform per-breath clas-sification of 5 of the most widely used ventilation modes in the USA with an average F1-score of 97.52%. We also show how our approach makes methodologic improvements over previous work and that it is highly robust to missing data caused by software/sensor error. |
Personal Mental Health Navigator: Harnessing the Power of Data, Personal Models, and Health Cybernetics to Promote Psychological Well-being | Traditionally, the regime of mental healthcare has followed an episodic psychotherapy model wherein patients seek care from a provider through a prescribed treatment plan developed over multiple provider visits. Recent advances in wearable and mobile technology have generated increased interest in digital mental healthcare that enables individuals to address episodic mental health symptoms. However, these efforts are typically reactive and symptom-focused and do not provide comprehensive, wrap-around, customized treatments that capture an individual's holistic mental health model as it unfolds over time. Recognizing that each individual is unique, we present the notion of Personalized Mental Health Navigation (MHN): a therapist-in-the-loop, cybernetic goal-based system that deploys a continuous cyclic loop of measurement, estimation, guidance, to steer the individual's mental health state towards a healthy zone. We outline the major components of MHN that is premised on the development of an individual's personal mental health state, holistically represented by a high-dimensional cover of multiple knowledge layers such as emotion, biological patterns, sociology, behavior, and cognition. We demonstrate the feasibility of the personalized MHN approach via a 12-month pilot case study for holistic stress management in college students and highlight an instance of a therapist-in-the-loop intervention using MHN for monitoring, estimating, and proactively addressing moderately severe depression over a sustained period of time. We believe MHN paves the way to transform mental healthcare from the current passive, episodic, reactive process (where individuals seek help to address symptoms that have already manifested) to a continuous and navigational paradigm that leverages a personalized model of the individual, promising to deliver timely interventions to individuals in a holistic manner. |
Fast and Reliable Evaluation of Adversarial Robustness with Minimum-Margin Attack | The AutoAttack (AA) has been the most reliable method to evaluate adversarial robustness when considerable computational resources are available. However, the high computational cost (e.g., 100 times more than that of the project gradient descent attack) makes AA infeasible for practitioners with limited computational resources, and also hinders applications of AA in the adversarial training (AT). In this paper, we propose a novel method, minimum-margin (MM) attack, to fast and reliably evaluate adversarial robustness. Compared with AA, our method achieves comparable performance but only costs 3% of the computational time in extensive experiments. The reliability of our method lies in that we evaluate the quality of adversarial examples using the margin between two targets that can precisely identify the most adversarial example. The computational efficiency of our method lies in an effective Sequential TArget Ranking Selection (STARS) method, ensuring that the cost of the MM attack is independent of the number of classes. The MM attack opens a new way for evaluating adversarial robustness and provides a feasible and reliable way to generate high-quality adversarial examples in AT. |
Improved Time Warp Edit Distance -- A Parallel Dynamic Program in Linear Memory | Edit Distance is a classic family of dynamic programming problems, among which Time Warp Edit Distance refines the problem with the notion of a metric and temporal elasticity. A novel Improved Time Warp Edit Distance algorithm that is both massively parallelizable and requiring only linear storage is presented. This method uses the procession of a three diagonal band to cover the original dynamic program space. Every element of the diagonal update can be computed in parallel. The core method is a feature of the TWED Longest Common Subsequence data dependence and is applicable to dynamic programs that share similar band subproblem structure. The algorithm has been implemented as a CUDA C library with Python bindings. Speedups for challenging problems are phenomenal. |
on the effectiveness of generative adversarial network on anomaly detection | Identifying anomalies refers to detecting samples that do not resemble the training data distribution. Many generative models have been used to find anomalies, and among them, generative adversarial network (GAN)-based approaches are currently very popular. GANs mainly rely on the rich contextual information of these models to identify the actual training distribution. Following this analogy, we suggested a new unsupervised model based on GANs --a combination of an autoencoder and a GAN. Further, a new scoring function was introduced to target anomalies where a linear combination of the internal representation of the discriminator and the generator's visual representation, plus the encoded representation of the autoencoder, come together to define the proposed anomaly score. The model was further evaluated on benchmark datasets such as SVHN, CIFAR10, and MNIST, as well as a public medical dataset of leukemia images. In all the experiments, our model outperformed its existing counterparts while slightly improving the inference time. |
EEG-based Subjects Identification based on Biometrics of Imagined Speech using EMD | When brain activity is translated into commands for real applications, the potential for human capacities augmentation is promising. In this paper, EMD is used to decompose EEG signals during Imagined Speech in order to use it as a biometric marker for creating a Biometric Recognition System. For each EEG channel, the most relevant Intrinsic Mode Functions (IMFs) are decided based on the Minkowski distance, and for each IMF 4 features are computed: Instantaneous and Teager energy distribution and Higuchi and Petrosian Fractal Dimension. To test the proposed method, a dataset with 20 subjects who imagined 30 repetitions of 5 words in Spanish, is used. Four classifiers are used for this task - random forest, SVM, naive Bayes, and k-NN - and their performances are compared. The accuracy obtained (up to 0.92 using Linear SVM) after 10-folds cross-validation suggest that the proposed method based on EMD can be valuable for creating EEG-based biometrics of imagined speech for Subjects identification. |
BLOB : A Probabilistic Model for Recommendation that Combines Organic and Bandit Signals | A common task for recommender systems is to build a pro le of the interests of a user from items in their browsing history and later to recommend items to the user from the same catalog. The users' behavior consists of two parts: the sequence of items that they viewed without intervention (the organic part) and the sequences of items recommended to them and their outcome (the bandit part). In this paper, we propose Bayesian Latent Organic Bandit model (BLOB), a probabilistic approach to combine the 'or-ganic' and 'bandit' signals in order to improve the estimation of recommendation quality. The bandit signal is valuable as it gives direct feedback of recommendation performance, but the signal quality is very uneven, as it is highly concentrated on the recommendations deemed optimal by the past version of the recom-mender system. In contrast, the organic signal is typically strong and covers most items, but is not always relevant to the recommendation task. In order to leverage the organic signal to e ciently learn the bandit signal in a Bayesian model we identify three fundamental types of distances, namely action-history, action-action and history-history distances. We implement a scalable approximation of the full model using variational auto-encoders and the local re-paramerization trick. We show using extensive simulation studies that our method out-performs or matches the value of both state-of-the-art organic-based recommendation algorithms, and of bandit-based methods (both value and policy-based) both in organic and bandit-rich environments. |
Online Continual Adaptation with Active Self-Training | Models trained with offline data often suffer from continual distribution shifts and expensive labeling in changing environments. This calls for a new online learning paradigm where the learner can continually adapt to changing environments with limited labels. In this paper, we propose a new online setting -- Online Active Continual Adaptation, where the learner aims to continually adapt to changing distributions using both unlabeled samples and active queries of limited labels. To this end, we propose Online Self-Adaptive Mirror Descent (OSAMD), which adopts an online teacher-student structure to enable online self-training from unlabeled data, and a margin-based criterion that decides whether to query the labels to track changing distributions. Theoretically, we show that, in the separable case, OSAMD has an $O({T}^{2/3})$ dynamic regret bound under mild assumptions, which is aligned with the $\Omega(T^{2/3})$ lower bound of online learning algorithms with full labels. In the general case, we show a regret bound of $O({T}^{2/3} + \alpha^* T)$, where $\alpha^*$ denotes the separability of domains and is usually small. Our theoretical results show that OSAMD can fast adapt to changing environments with active queries. Empirically, we demonstrate that OSAMD achieves favorable regrets under changing environments with limited labels on both simulated and real-world data, which corroborates our theoretical findings. |
Convex Analysis of the Mean Field Langevin Dynamics | As an example of the nonlinear Fokker-Planck equation, the mean field Langevin dynamics recently attracts attention due to its connection to (noisy) gradient descent on infinitely wide neural networks in the mean field regime, and hence the convergence property of the dynamics is of great theoretical interest. In this work, we give a concise and self-contained convergence rate analysis of the mean field Langevin dynamics with respect to the (regularized) objective function in both continuous and discrete time settings. The key ingredient of our proof is a proximal Gibbs distribution $p_q$ associated with the dynamics, which, in combination with techniques in [Vempala and Wibisono (2019)], allows us to develop a simple convergence theory parallel to classical results in convex optimization. Furthermore, we reveal that $p_q$ connects to the duality gap in the empirical risk minimization setting, which enables efficient empirical evaluation of the algorithm convergence. |
Enhanced Self-Organizing Map Solution for the Traveling Salesman Problem | Using an enhanced Self-Organizing Map method, we provided suboptimal solutions to the Traveling Salesman Problem. Besides, we employed hyperparameter tuning to identify the most critical features in the algorithm. All improvements in the benchmark work brought consistent results and may inspire future efforts to improve this algorithm and apply it to different problems. |
Detecting the Starting Frame of Actions in Video | In this work, we address the problem of precisely localizing key frames of an action, for example, the precise time that a pitcher releases a baseball, or the precise time that a crowd begins to applaud. Key frame localization is a largely overlooked and important action-recognition problem, for example in the field of neuroscience, in which we would like to understand the neural activity that produces the start of a bout of an action. To address this problem, we introduce a novel structured loss function that properly weights the types of errors that matter in such applications: it more heavily penalizes extra and missed action start detections over small misalignments. Our structured loss is based on the best matching between predicted and labeled action starts. We train recurrent neural networks (RNNs) to minimize differentiable approximations of this loss. To evaluate these methods, we introduce the Mouse Reach Dataset, a large, annotated video dataset of mice performing a sequence of actions. The dataset was collected and labeled by experts for the purpose of neuroscience research. On this dataset, we demonstrate that our method outperforms related approaches and baseline methods using an unstructured loss. |
When Do Differences Matter? On-Line Feature Extraction Through Cognitive Economy | For an intelligent agent to be truly autonomous, it must be able to adapt its representation to the requirements of its task as it interacts with the world. Most current approaches to on-line feature extraction are ad hoc; in contrast, this paper presents an algorithm that bases judgments of state compatibility and state-space abstraction on principled criteria derived from the psychological principle of cognitive economy. The algorithm incorporates an active form of Q-learning, and partitions continuous state-spaces by merging and splitting Voronoi regions. The experiments illustrate a new methodology for testing and comparing representations by means of learning curves. Results from the puck-on-a-hill task demonstrate the algorithm's ability to learn effective representations, superior to those produced by some other, well-known, methods. |
Gaze-Sensing LEDs for Head Mounted Displays | We introduce a new gaze tracker for Head Mounted Displays (HMDs). We modify two off-the-shelf HMDs to be gaze-aware using Light Emitting Diodes (LEDs). Our key contribution is to exploit the sensing capability of LEDs to create low-power gaze tracker for virtual reality (VR) applications. This yields a simple approach using minimal hardware to achieve good accuracy and low latency using light-weight supervised Gaussian Process Regression (GPR) running on a mobile device. With our hardware, we show that Minkowski distance measure based GPR implementation outperforms the commonly used radial basis function-based support vector regression (SVR) without the need to precisely determine free parameters. We show that our gaze estimation method does not require complex dimension reduction techniques, feature extraction, or distortion corrections due to off-axis optical paths. We demonstrate two complete HMD prototypes with a sample eye-tracked application, and report on a series of subjective tests using our prototypes. |
Accounting for Model Uncertainty in Algorithmic Discrimination | Traditional approaches to ensure group fairness in algorithmic decision making aim to equalize ``total'' error rates for different subgroups in the population. In contrast, we argue that the fairness approaches should instead focus only on equalizing errors arising due to model uncertainty (a.k.a epistemic uncertainty), caused due to lack of knowledge about the best model or due to lack of data. In other words, our proposal calls for ignoring the errors that occur due to uncertainty inherent in the data, i.e., aleatoric uncertainty. We draw a connection between predictive multiplicity and model uncertainty and argue that the techniques from predictive multiplicity could be used to identify errors made due to model uncertainty. We propose scalable convex proxies to come up with classifiers that exhibit predictive multiplicity and empirically show that our methods are comparable in performance and up to four orders of magnitude faster than the current state-of-the-art. We further propose methods to achieve our goal of equalizing group error rates arising due to model uncertainty in algorithmic decision making and demonstrate the effectiveness of these methods using synthetic and real-world datasets. |
Error bounds in estimating the out-of-sample prediction error using leave-one-out cross validation in high-dimensions | We study the problem of out-of-sample risk estimation in the high dimensional regime where both the sample size $n$ and number of features $p$ are large, and $n/p$ can be less than one. Extensive empirical evidence confirms the accuracy of leave-one-out cross validation (LO) for out-of-sample risk estimation. Yet, a unifying theoretical evaluation of the accuracy of LO in high-dimensional problems has remained an open problem. This paper aims to fill this gap for penalized regression in the generalized linear family. With minor assumptions about the data generating process, and without any sparsity assumptions on the regression coefficients, our theoretical analysis obtains finite sample upper bounds on the expected squared error of LO in estimating the out-of-sample error. Our bounds show that the error goes to zero as $n,p \rightarrow \infty$, even when the dimension $p$ of the feature vectors is comparable with or greater than the sample size $n$. One technical advantage of the theory is that it can be used to clarify and connect some results from the recent literature on scalable approximate LO. |
Bayesian Image Classification with Deep Convolutional Gaussian Processes | In decision-making systems, it is important to have classifiers that have calibrated uncertainties, with an optimisation objective that can be used for automated model selection and training. Gaussian processes (GPs) provide uncertainty estimates and a marginal likelihood objective, but their weak inductive biases lead to inferior accuracy. This has limited their applicability in certain tasks (e.g. image classification). We propose a translation-insensitive convolutional kernel, which relaxes the translation invariance constraint imposed by previous convolutional GPs. We show how we can use the marginal likelihood to learn the degree of insensitivity. We also reformulate GP image-to-image convolutional mappings as multi-output GPs, leading to deep convolutional GPs. We show experimentally that our new kernel improves performance in both single-layer and deep models. We also demonstrate that our fully Bayesian approach improves on dropout-based Bayesian deep learning methods in terms of uncertainty and marginal likelihood estimates. |
FIND: Human-in-the-Loop Debugging Deep Text Classifiers | Since obtaining a perfect training dataset (i.e., a dataset which is considerably large, unbiased, and well-representative of unseen cases) is hardly possible, many real-world text classifiers are trained on the available, yet imperfect, datasets. These classifiers are thus likely to have undesirable properties. For instance, they may have biases against some sub-populations or may not work effectively in the wild due to overfitting. In this paper, we propose FIND -- a framework which enables humans to debug deep learning text classifiers by disabling irrelevant hidden features. Experiments show that by using FIND, humans can improve CNN text classifiers which were trained under different types of imperfect datasets (including datasets with biases and datasets with dissimilar train-test distributions). |
Compositional Verification for Autonomous Systems with Deep Learning Components | As autonomy becomes prevalent in many applications, ranging from recommendation systems to fully autonomous vehicles, there is an increased need to provide safety guarantees for such systems. The problem is difficult, as these are large, complex systems which operate in uncertain environments, requiring data-driven machine-learning components. However, learning techniques such as Deep Neural Networks, widely used today, are inherently unpredictable and lack the theoretical foundations to provide strong assurance guarantees. We present a compositional approach for the scalable, formal verification of autonomous systems that contain Deep Neural Network components. The approach uses assume-guarantee reasoning whereby {\em contracts}, encoding the input-output behavior of individual components, allow the designer to model and incorporate the behavior of the learning-enabled components working side-by-side with the other components. We illustrate the approach on an example taken from the autonomous vehicles domain. |
A Fast Quartet Tree Heuristic for Hierarchical Clustering | The Minimum Quartet Tree Cost problem is to construct an optimal weight tree from the $3{n \choose 4}$ weighted quartet topologies on $n$ objects, where optimality means that the summed weight of the embedded quartet topologies is optimal (so it can be the case that the optimal tree embeds all quartets as nonoptimal topologies). We present a Monte Carlo heuristic, based on randomized hill climbing, for approximating the optimal weight tree, given the quartet topology weights. The method repeatedly transforms a dendrogram, with all objects involved as leaves, achieving a monotonic approximation to the exact single globally optimal tree. The problem and the solution heuristic has been extensively used for general hierarchical clustering of nontree-like (non-phylogeny) data in various domains and across domains with heterogeneous data. We also present a greatly improved heuristic, reducing the running time by a factor of order a thousand to ten thousand. All this is implemented and available, as part of the CompLearn package. We compare performance and running time of the original and improved versions with those of UPGMA, BioNJ, and NJ, as implemented in the SplitsTree package on genomic data for which the latter are optimized. Keywords: Data and knowledge visualization, Pattern matching--Clustering--Algorithms/Similarity measures, Hierarchical clustering, Global optimization, Quartet tree, Randomized hill-climbing, |
Automatic Detection of Satire in Bangla Documents: A CNN Approach Based on Hybrid Feature Extraction Model | Widespread of satirical news in online communities is an ongoing trend. The nature of satires is so inherently ambiguous that sometimes it's too hard even for humans to understand whether it's actually satire or not. So, research interest has grown in this field. The purpose of this research is to detect Bangla satirical news spread in online news portals as well as social media. In this paper, we propose a hybrid technique for extracting features from text documents combining Word2Vec and TF-IDF. Using our proposed feature extraction technique, with standard CNN architecture we could detect whether a Bangla text document is satire or not with an accuracy of more than 96%. |
DeSIGN: Design Inspiration from Generative Networks | Can an algorithm create original and compelling fashion designs to serve as an inspirational assistant? To help answer this question, we design and investigate different image generation models associated with different loss functions to boost creativity in fashion generation. The dimensions of our explorations include: (i) different Generative Adversarial Networks architectures that start from noise vectors to generate fashion items, (ii) novel loss functions that encourage novelty, inspired from Sharma-Mittal divergence, a generalized mutual information measure for the widely used relative entropies such as Kullback-Leibler, and (iii) a generation process following the key elements of fashion design (disentangling shape and texture components). A key challenge of this study is the evaluation of generated designs and the retrieval of best ones, hence we put together an evaluation protocol associating automatic metrics and human experimental studies that we hope will help ease future research. We show that our proposed creativity criterion yield better overall appreciation than the one employed in Creative Adversarial Networks. In the end, about 61% of our images are thought to be created by human designers rather than by a computer while also being considered original per our human subject experiments, and our proposed loss scores the highest compared to existing losses in both novelty and likability. |
RPM-Net: Recurrent Prediction of Motion and Parts from Point Cloud | We introduce RPM-Net, a deep learning-based approach which simultaneously infers movable parts and hallucinates their motions from a single, un-segmented, and possibly partial, 3D point cloud shape. RPM-Net is a novel Recurrent Neural Network (RNN), composed of an encoder-decoder pair with interleaved Long Short-Term Memory (LSTM) components, which together predict a temporal sequence of pointwise displacements for the input point cloud. At the same time, the displacements allow the network to learn movable parts, resulting in a motion-based shape segmentation. Recursive applications of RPM-Net on the obtained parts can predict finer-level part motions, resulting in a hierarchical object segmentation. Furthermore, we develop a separate network to estimate part mobilities, e.g., per-part motion parameters, from the segmented motion sequence. Both networks learn deep predictive models from a training set that exemplifies a variety of mobilities for diverse objects. We show results of simultaneous motion and part predictions from synthetic and real scans of 3D objects exhibiting a variety of part mobilities, possibly involving multiple movable parts. |
Comparing EM with GD in Mixture Models of Two Components | The expectation-maximization (EM) algorithm has been widely used in minimizing the negative log likelihood (also known as cross entropy) of mixture models. However, little is understood about the goodness of the fixed points it converges to. In this paper, we study the regions where one component is missing in two-component mixture models, which we call one-cluster regions. We analyze the propensity of such regions to trap EM and gradient descent (GD) for mixtures of two Gaussians and mixtures of two Bernoullis. In the case of Gaussian mixtures, EM escapes one-cluster regions exponentially fast, while GD escapes them linearly fast. In the case of mixtures of Bernoullis, we find that there exist one-cluster regions that are stable for GD and therefore trap GD, but those regions are unstable for EM, allowing EM to escape. Those regions are local minima that appear universally in experiments and can be arbitrarily bad. This work implies that EM is less likely than GD to converge to certain bad local optima in mixture models. |
Stochastic Generalized Adversarial Label Learning | The usage of machine learning models has grown substantially and is spreading into several application domains. A common need in using machine learning models is collecting the data required to train these models. In some cases, labeling a massive dataset can be a crippling bottleneck, so there is need to develop models that work when training labels for large amounts of data are not easily obtained. A possible solution is weak supervision, which uses noisy labels that are easily obtained from multiple sources. The challenge is how best to combine these noisy labels and train a model to perform well given a task. In this paper, we propose stochastic generalized adversarial label learning (Stoch-GALL), a framework for training machine learning models that perform well when noisy and possibly correlated labels are provided. Our framework allows users to provide different weak labels and multiple constraints on these labels. Our model then attempts to learn parameters for the data by solving a non-zero sum game optimization. The game is between an adversary that chooses labels for the data and a model that minimizes the error made by the adversarial labels. We test our method on three datasets by training convolutional neural network models that learn to classify image objects with limited access to training labels. Our approach is able to learn even in settings where the weak supervision confounds state-of-the-art weakly supervised learning methods. The results of our experiments demonstrate the applicability of this approach to general classification tasks. |
A Progressive Conditional Generative Adversarial Network for Generating Dense and Colored 3D Point Clouds | In this paper, we introduce a novel conditional generative adversarial network that creates dense 3D point clouds, with color, for assorted classes of objects in an unsupervised manner. To overcome the difficulty of capturing intricate details at high resolutions, we propose a point transformer that progressively grows the network through the use of graph convolutions. The network is composed of a leaf output layer and an initial set of branches. Every training iteration evolves a point vector into a point cloud of increasing resolution. After a fixed number of iterations, the number of branches is increased by replicating the last branch. Experimental results show that our network is capable of learning and mimicking a 3D data distribution, and produces colored point clouds with fine details at multiple resolutions. |
Recurrent Feedback Improves Feedforward Representations in Deep Neural Networks | The abundant recurrent horizontal and feedback connections in the primate visual cortex are thought to play an important role in bringing global and semantic contextual information to early visual areas during perceptual inference, helping to resolve local ambiguity and fill in missing details. In this study, we find that introducing feedback loops and horizontal recurrent connections to a deep convolution neural network (VGG16) allows the network to become more robust against noise and occlusion during inference, even in the initial feedforward pass. This suggests that recurrent feedback and contextual modulation transform the feedforward representations of the network in a meaningful and interesting way. We study the population codes of neurons in the network, before and after learning with feedback, and find that learning with feedback yielded an increase in discriminability (measured by d-prime) between the different object classes in the population codes of the neurons in the feedforward path, even at the earliest layer that receives feedback. We find that recurrent feedback, by injecting top-down semantic meaning to the population activities, helps the network learn better feedforward paths to robustly map noisy image patches to the latent representations corresponding to important visual concepts of each object class, resulting in greater robustness of the network against noises and occlusion as well as better fine-grained recognition. |
Meta-Reinforcement Learning with Self-Modifying Networks | Deep Reinforcement Learning has demonstrated the potential of neural networks tuned with gradient descent for solving complex tasks in well-delimited environments. However, these neural systems are slow learners producing specialised agents with no mechanism to continue learning beyond their training curriculum. On the contrary, biological synaptic plasticity is persistent and manifold, and has been hypothesised to play a key role in executive functions such as working memory and cognitive flexibility, potentially supporting more efficient and generic learning abilities. Inspired by this, we propose to build networks with dynamic weights, able to continually perform self-reflexive modification as a function of their current synaptic state and action-reward feedback, rather than a fixed network configuration. The resulting model, MetODS (for Meta-Optimized Dynamical Synapses) is a broadly applicable meta-reinforcement learning system able to learn efficient and powerful control rules in the agent policy space. A single layer with dynamic synapses can perform one-shot learning, generalize navigation principles to unseen environments and demonstrate a strong ability to learn adaptive motor policies, comparing favourably with previous meta-reinforcement learning approaches. |
Graph similarity learning for change-point detection in dynamic networks | Dynamic networks are ubiquitous for modelling sequential graph-structured data, e.g., brain connectome, population flows and messages exchanges. In this work, we consider dynamic networks that are temporal sequences of graph snapshots, and aim at detecting abrupt changes in their structure. This task is often termed network change-point detection and has numerous applications, such as fraud detection or physical motion monitoring. Leveraging a graph neural network model, we design a method to perform online network change-point detection that can adapt to the specific network domain and localise changes with no delay. The main novelty of our method is to use a siamese graph neural network architecture for learning a data-driven graph similarity function, which allows to effectively compare the current graph and its recent history. Importantly, our method does not require prior knowledge on the network generative distribution and is agnostic to the type of change-points; moreover, it can be applied to a large variety of networks, that include for instance edge weights and node attributes. We show on synthetic and real data that our method enjoys a number of benefits: it is able to learn an adequate graph similarity function for performing online network change-point detection in diverse types of change-point settings, and requires a shorter data history to detect changes than most existing state-of-the-art baselines. |
Online Deep Learning based on Auto-Encoder | Online learning is an important technical means for sketching massive real-time and high-speed data. Although this direction has attracted intensive attention, most of the literature in this area ignore the following three issues: (1) they think little of the underlying abstract hierarchical latent information existing in examples, even if extracting these abstract hierarchical latent representations is useful to better predict the class labels of examples; (2) the idea of preassigned model on unseen datapoints is not suitable for modeling streaming data with evolving probability distribution. This challenge is referred as model flexibility. And so, with this in minds, the online deep learning model we need to design should have a variable underlying structure; (3) moreover, it is of utmost importance to fusion these abstract hierarchical latent representations to achieve better classification performance, and we should give different weights to different levels of implicit representation information when dealing with the data streaming where the data distribution changes. To address these issues, we propose a two-phase Online Deep Learning based on Auto-Encoder (ODLAE). Based on auto-encoder, considering reconstruction loss, we extract abstract hierarchical latent representations of instances; Based on predictive loss, we devise two fusion strategies: the output-level fusion strategy, which is obtained by fusing the classification results of encoder each hidden layer; and feature-level fusion strategy, which is leveraged self-attention mechanism to fusion every hidden layer output. Finally, in order to improve the robustness of the algorithm, we also try to utilize the denoising auto-encoder to yield hierarchical latent representations. Experimental results on different datasets are presented to verify the validity of our proposed algorithm (ODLAE) outperforms several baselines. |
Efficient Few-Shot Fine-Tuning for Opinion Summarization | Abstractive summarization models are typically pre-trained on large amounts of generic texts, then fine-tuned on tens or hundreds of thousands of annotated samples. However, in opinion summarization, large annotated datasets of reviews paired with reference summaries are not available and would be expensive to create. This calls for fine-tuning methods robust to overfitting on small datasets. In addition, generically pre-trained models are often not accustomed to the specifics of customer reviews and, after fine-tuning, yield summaries with disfluencies and semantic mistakes. To address these problems, we utilize an efficient few-shot method based on adapters which, as we show, can easily store in-domain knowledge. Instead of fine-tuning the entire model, we add adapters and pre-train them in a task-specific way on a large corpus of unannotated customer reviews, using held-out reviews as pseudo summaries. Then, fine-tune the adapters on the small available human-annotated dataset. We show that this self-supervised adapter pre-training improves summary quality over standard fine-tuning by 2.0 and 1.3 ROUGE-L points on the Amazon and Yelp datasets, respectively. Finally, for summary personalization, we condition on aspect keyword queries, automatically created from generic datasets. In the same vein, we pre-train the adapters in a query-based manner on customer reviews and then fine-tune them on annotated datasets. This results in better-organized summary content reflected in improved coherence and fewer redundancies. |
Dynamic Edge Weights in Graph Neural Networks for 3D Object Detection | A robust and accurate 3D detection system is an integral part of autonomous vehicles. Traditionally, a majority of 3D object detection algorithms focus on processing 3D point clouds using voxel grids or bird's eye view (BEV). Recent works, however, demonstrate the utilization of the graph neural network (GNN) as a promising approach to 3D object detection. In this work, we propose an attention based feature aggregation technique in GNN for detecting objects in LiDAR scan. We first employ a distance-aware down-sampling scheme that not only enhances the algorithmic performance but also retains maximum geometric features of objects even if they lie far from the sensor. In each layer of the GNN, apart from the linear transformation which maps the per node input features to the corresponding higher level features, a per node masked attention by specifying different weights to different nodes in its first ring neighborhood is also performed. The masked attention implicitly accounts for the underlying neighborhood graph structure of every node and also eliminates the need of costly matrix operations thereby improving the detection accuracy without compromising the performance. The experiments on KITTI dataset show that our method yields comparable results for 3D object detection. |
Deep Contextualized Acoustic Representations For Semi-Supervised Speech Recognition | We propose a novel approach to semi-supervised automatic speech recognition (ASR). We first exploit a large amount of unlabeled audio data via representation learning, where we reconstruct a temporal slice of filterbank features from past and future context frames. The resulting deep contextualized acoustic representations (DeCoAR) are then used to train a CTC-based end-to-end ASR system using a smaller amount of labeled audio data. In our experiments, we show that systems trained on DeCoAR consistently outperform ones trained on conventional filterbank features, giving 42% and 19% relative improvement over the baseline on WSJ eval92 and LibriSpeech test-clean, respectively. Our approach can drastically reduce the amount of labeled data required; unsupervised training on LibriSpeech then supervision with 100 hours of labeled data achieves performance on par with training on all 960 hours directly. Pre-trained models and code will be released online. |
Neural Networks Enhancement with Logical Knowledge | In the recent past, there has been a growing interest in Neural-Symbolic Integration frameworks, i.e., hybrid systems that integrate connectionist and symbolic approaches to obtain the best of both worlds. In a previous work, we proposed KENN (Knowledge Enhanced Neural Networks), a Neural-Symbolic architecture that injects prior logical knowledge into a neural network by adding a new final layer which modifies the initial predictions accordingly to the knowledge. Among the advantages of this strategy, there is the inclusion of clause weights, learnable parameters that represent the strength of the clauses, meaning that the model can learn the impact of each clause on the final predictions. As a special case, if the training data contradicts a constraint, KENN learns to ignore it, making the system robust to the presence of wrong knowledge. In this paper, we propose an extension of KENN for relational data. To evaluate this new extension, we tested it with different learning configurations on Citeseer, a standard dataset for Collective Classification. The results show that KENN is capable of increasing the performances of the underlying neural network even in the presence relational data, outperforming other two notable methods that combine learning with logic. |
Deep Networks with Shape Priors for Nucleus Detection | Detection of cell nuclei in microscopic images is a challenging research topic, because of limitations in cellular image quality and diversity of nuclear morphology, i.e. varying nuclei shapes, sizes, and overlaps between multiple cell nuclei. This has been a topic of enduring interest with promising recent success shown by deep learning methods. These methods train for example convolutional neural networks (CNNs) with a training set of input images and known, labeled nuclei locations. Many of these methods are supplemented by spatial or morphological processing. We develop a new approach that we call Shape Priors with Convolutional Neural Networks (SP-CNN) to perform significantly enhanced nuclei detection. A set of canonical shapes is prepared with the help of a domain expert. Subsequently, we present a new network structure that can incorporate `expected behavior' of nucleus shapes via two components: {\em learnable} layers that perform the nucleus detection and a {\em fixed} processing part that guides the learning with prior information. Analytically, we formulate a new regularization term that is targeted at penalizing false positives while simultaneously encouraging detection inside cell nucleus boundary. Experimental results on a challenging dataset reveal that SP-CNN is competitive with or outperforms several state-of-the-art methods. |
Talking-Heads Attention | We introduce "talking-heads attention" - a variation on multi-head attention which includes linearprojections across the attention-heads dimension, immediately before and after the softmax operation.While inserting only a small number of additional parameters and a moderate amount of additionalcomputation, talking-heads attention leads to better perplexities on masked language modeling tasks, aswell as better quality when transfer-learning to language comprehension and question answering tasks. |
Robust Semi-Supervised Graph Classifier Learning with Negative Edge Weights | In a semi-supervised learning scenario, (possibly noisy) partially observed labels are used as input to train a classifier, in order to assign labels to unclassified samples. In this paper, we study this classifier learning problem from a graph signal processing (GSP) perspective. Specifically, by viewing a binary classifier as a piecewise constant graph-signal in a high-dimensional feature space, we cast classifier learning as a signal restoration problem via a classical maximum a posteriori (MAP) formulation. Unlike previous graph-signal restoration works, we consider in addition edges with negative weights that signify anti-correlation between samples. One unfortunate consequence is that the graph Laplacian matrix $\mathbf{L}$ can be indefinite, and previously proposed graph-signal smoothness prior $\mathbf{x}^T \mathbf{L} \mathbf{x}$ for candidate signal $\mathbf{x}$ can lead to pathological solutions. In response, we derive an optimal perturbation matrix $\boldsymbol{\Delta}$ - based on a fast lower-bound computation of the minimum eigenvalue of $\mathbf{L}$ via a novel application of the Haynsworth inertia additivity formula---so that $\mathbf{L} + \boldsymbol{\Delta}$ is positive semi-definite, resulting in a stable signal prior. Further, instead of forcing a hard binary decision for each sample, we define the notion of generalized smoothness on graph that promotes ambiguity in the classifier signal. Finally, we propose an algorithm based on iterative reweighted least squares (IRLS) that solves the posed MAP problem efficiently. Extensive simulation results show that our proposed algorithm outperforms both SVM variants and graph-based classifiers using positive-edge graphs noticeably. |
In-Distribution Interpretability for Challenging Modalities | It is widely recognized that the predictions of deep neural networks are difficult to parse relative to simpler approaches. However, the development of methods to investigate the mode of operation of such models has advanced rapidly in the past few years. Recent work introduced an intuitive framework which utilizes generative models to improve on the meaningfulness of such explanations. In this work, we display the flexibility of this method to interpret diverse and challenging modalities: music and physical simulations of urban environments. |
Empowering parameter-efficient transfer learning by recognizing the kernel structure in self-attention | The massive amount of trainable parameters in the pre-trained language models (PLMs) makes them hard to be deployed to multiple downstream tasks. To address this issue, parameter-efficient transfer learning methods have been proposed to tune only a few parameters during fine-tuning while freezing the rest. This paper looks at existing methods along this line through the \textit{kernel lens}. Motivated by the connection between self-attention in transformer-based PLMs and kernel learning, we propose \textit{kernel-wise adapters}, namely \textit{Kernel-mix}, that utilize the kernel structure in self-attention to guide the assignment of the tunable parameters. These adapters use guidelines found in classical kernel learning and enable separate parameter tuning for each attention head. Our empirical results, over a diverse set of natural language generation and understanding tasks, show that our proposed adapters can attain or improve the strong performance of existing baselines. |
Bridging Model-based Safety and Model-free Reinforcement Learning through System Identification of Low Dimensional Linear Models | Bridging model-based safety and model-free reinforcement learning (RL) for dynamic robots is appealing since model-based methods are able to provide formal safety guarantees, while RL-based methods are able to exploit the robot agility by learning from the full-order system dynamics. However, current approaches to tackle this problem are mostly restricted to simple systems. In this paper, we propose a new method to combine model-based safety with model-free reinforcement learning by explicitly finding a low-dimensional model of the system controlled by a RL policy and applying stability and safety guarantees on that simple model. We use a complex bipedal robot Cassie, which is a high dimensional nonlinear system with hybrid dynamics and underactuation, and its RL-based walking controller as an example. We show that a low-dimensional dynamical model is sufficient to capture the dynamics of the closed-loop system. We demonstrate that this model is linear, asymptotically stable, and is decoupled across control input in all dimensions. We further exemplify that such linearity exists even when using different RL control policies. Such results point out an interesting direction to understand the relationship between RL and optimal control: whether RL tends to linearize the nonlinear system during training in some cases. Furthermore, we illustrate that the found linear model is able to provide guarantees by safety-critical optimal control framework, e.g., Model Predictive Control with Control Barrier Functions, on an example of autonomous navigation using Cassie while taking advantage of the agility provided by the RL-based controller. |
SingAug: Data Augmentation for Singing Voice Synthesis with Cycle-consistent Training Strategy | Deep learning based singing voice synthesis (SVS) systems have been demonstrated to flexibly generate singing with better qualities, compared to conventional statistical parametric based methods. However, neural systems are generally data-hungry and have difficulty to reach reasonable singing quality with limited public available training data. In this work, we explore different data augmentation methods to boost the training of SVS systems, including several strategies customized to SVS based on pitch augmentation and mix-up augmentation. To further stabilize the training, we introduce the cycle-consistent training strategy. Extensive experiments on two public singing databases demonstrate that our proposed augmentation methods and the stabilizing training strategy can significantly improve the performance on both objective and subjective evaluations. |
Explanations for Occluded Images | Existing algorithms for explaining the output of image classifiers perform poorly on inputs where the object of interest is partially occluded. We present a novel, black-box algorithm for computing explanations that uses a principled approach based on causal theory. We have implemented the method in the DEEPCOVER tool. We obtain explanations that are much more accurate than those generated by the existing explanation tools on images with occlusions and observe a level of performance comparable to the state of the art when explaining images without occlusions. |
Towards better social crisis data with HERMES: Hybrid sensing for EmeRgency ManagEment System | People involved in mass emergencies increasingly publish information-rich contents in online social networks (OSNs), thus acting as a distributed and resilient network of human sensors. In this work, we present HERMES, a system designed to enrich the information spontaneously disclosed by OSN users in the aftermath of disasters. HERMES leverages a mixed data collection strategy, called hybrid crowdsensing, and state-of-the-art AI techniques. Evaluated in real-world emergencies, HERMES proved to increase: (i) the amount of the available damage information; (ii) the density (up to 7x) and the variety (up to 18x) of the retrieved geographic information; (iii) the geographic coverage (up to 30%) and granularity. |
Proceedings of the 2016 ICML Workshop on #Data4Good: Machine Learning in Social Good Applications | This is the Proceedings of the ICML Workshop on #Data4Good: Machine Learning in Social Good Applications, which was held on June 24, 2016 in New York. |
Using Computational Intelligence for solving the Ornstein-Zernike equation | The main goal of this thesis is to provide an exploration of the use of computational intelligence techniques to study the numerical solution of the Ornstein-Zernike equation for simple liquids. In particular, a continuous model of the hard sphere fluid is studied. There are two main proposals in this contribution. First, the use of neural networks as a way to parametrize closure relation when solving the Ornstein-Zernike equation. It is explicitly shown that in the case of the hard sphere fluid, the neural network approach seems to reduce to the so-called Hypernetted Chain closure. For the second proposal, we explore the fact that if more physical information is incorporated into the theoretical formalism, a better estimate can be obtained with the use of evolutionary optimization techniques. When choosing the modified Verlet closure relation, and leaving a couple of free parameters to be adjusted, the results are as good as those obtained from molecular simulations. The thesis is then closed with a brief summary of the main findings and outlooks on different ways to improve the proposals presented here. |
Diverse, Global and Amortised Counterfactual Explanations for Uncertainty Estimates | To interpret uncertainty estimates from differentiable probabilistic models, recent work has proposed generating a single Counterfactual Latent Uncertainty Explanation (CLUE) for a given data point where the model is uncertain, identifying a single, on-manifold change to the input such that the model becomes more certain in its prediction. We broaden the exploration to examine $\delta$-CLUE, the set of potential CLUEs within a $\delta$ ball of the original input in latent space. We study the diversity of such sets and find that many CLUEs are redundant; as such, we propose DIVerse CLUE ($\nabla$-CLUE), a set of CLUEs which each propose a distinct explanation as to how one can decrease the uncertainty associated with an input. We then further propose GLobal AMortised CLUE (GLAM-CLUE), a distinct and novel method which learns amortised mappings on specific groups of uncertain inputs, taking them and efficiently transforming them in a single function call into inputs for which a model will be certain. Our experiments show that $\delta$-CLUE, $\nabla$-CLUE, and GLAM-CLUE all address shortcomings of CLUE and provide beneficial explanations of uncertainty estimates to practitioners. |
Lower-Bounded Proper Losses for Weakly Supervised Classification | This paper discusses the problem of weakly supervised classification, in which instances are given weak labels that are produced by some label-corruption process. The goal is to derive conditions under which loss functions for weak-label learning are proper and lower-bounded -- two essential requirements for the losses used in class-probability estimation. To this end, we derive a representation theorem for proper losses in supervised learning, which dualizes the Savage representation. We use this theorem to characterize proper weak-label losses and find a condition for them to be lower-bounded. From these theoretical findings, we derive a novel regularization scheme called generalized logit squeezing, which makes any proper weak-label loss bounded from below, without losing properness. Furthermore, we experimentally demonstrate the effectiveness of our proposed approach, as compared to improper or unbounded losses. The results highlight the importance of properness and lower-boundedness. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.