title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Global Convergence of Block Coordinate Descent in Deep Learning | Deep learning has aroused extensive attention due to its great empirical
success. The efficiency of the block coordinate descent (BCD) methods has been
recently demonstrated in deep neural network (DNN) training. However,
theoretical studies on their convergence properties are limited due to the
highly nonconvex nature of DNN training. In this paper, we aim at providing a
general methodology for provable convergence guarantees for this type of
methods. In particular, for most of the commonly used DNN training models
involving both two- and three-splitting schemes, we establish the global
convergence to a critical point at a rate of ${\cal O}(1/k)$, where $k$ is the
number of iterations. The results extend to general loss functions which have
Lipschitz continuous gradients and deep residual networks (ResNets). Our key
development adds several new elements to the Kurdyka-{\L}ojasiewicz inequality
framework that enables us to carry out the global convergence analysis of BCD
in the general scenario of deep learning.
|
WRPN & Apprentice: Methods for Training and Inference using
Low-Precision Numerics | Today's high performance deep learning architectures involve large models
with numerous parameters. Low precision numerics has emerged as a popular
technique to reduce both the compute and memory requirements of these large
models. However, lowering precision often leads to accuracy degradation. We
describe three schemes whereby one can both train and do efficient inference
using low precision numerics without hurting accuracy. Finally, we describe an
efficient hardware accelerator that can take advantage of the proposed low
precision numerics.
|
DRUNET: A Dilated-Residual U-Net Deep Learning Network to Digitally
Stain Optic Nerve Head Tissues in Optical Coherence Tomography Images | Given that the neural and connective tissues of the optic nerve head (ONH)
exhibit complex morphological changes with the development and progression of
glaucoma, their simultaneous isolation from optical coherence tomography (OCT)
images may be of great interest for the clinical diagnosis and management of
this pathology. A deep learning algorithm was designed and trained to digitally
stain (i.e. highlight) 6 ONH tissue layers by capturing both the local (tissue
texture) and contextual information (spatial arrangement of tissues). The
overall dice coefficient (mean of all tissues) was $0.91 \pm 0.05$ when
assessed against manual segmentations performed by an expert observer. We offer
here a robust segmentation framework that could be extended for the automated
parametric study of the ONH tissues.
|
Distance Measure Machines | This paper presents a distance-based discriminative framework for learning
with probability distributions. Instead of using kernel mean embeddings or
generalized radial basis kernels, we introduce embeddings based on
dissimilarity of distributions to some reference distributions denoted as
templates. Our framework extends the theory of similarity of Balcan et al.
(2008) to the population distribution case and we show that, for some learning
problems, some dissimilarity on distribution achieves low-error linear decision
functions with high probability. Our key result is to prove that the theory
also holds for empirical distributions. Algorithmically, the proposed approach
consists in computing a mapping based on pairwise dissimilarity where learning
a linear decision function is amenable. Our experimental results show that the
Wasserstein distance embedding performs better than kernel mean embeddings and
computing Wasserstein distance is far more tractable than estimating pairwise
Kullback-Leibler divergence of empirical distributions.
|
Model-Based Clustering and Classification of Functional Data | The problem of complex data analysis is a central topic of modern statistical
science and learning systems and is becoming of broader interest with the
increasing prevalence of high-dimensional data. The challenge is to develop
statistical models and autonomous algorithms that are able to acquire knowledge
from raw data for exploratory analysis, which can be achieved through
clustering techniques or to make predictions of future data via classification
(i.e., discriminant analysis) techniques. Latent data models, including mixture
model-based approaches are one of the most popular and successful approaches in
both the unsupervised context (i.e., clustering) and the supervised one (i.e,
classification or discrimination). Although traditionally tools of multivariate
analysis, they are growing in popularity when considered in the framework of
functional data analysis (FDA). FDA is the data analysis paradigm in which the
individual data units are functions (e.g., curves, surfaces), rather than
simple vectors. In many areas of application, the analyzed data are indeed
often available in the form of discretized values of functions or curves (e.g.,
time series, waveforms) and surfaces (e.g., 2d-images, spatio-temporal data).
This functional aspect of the data adds additional difficulties compared to the
case of a classical multivariate (non-functional) data analysis. We review and
present approaches for model-based clustering and classification of functional
data. We derive well-established statistical models along with efficient
algorithmic tools to address problems regarding the clustering and the
classification of these high-dimensional data, including their heterogeneity,
missing information, and dynamical hidden structure. The presented models and
algorithms are illustrated on real-world functional data analysis problems from
several application area.
|
Minimax rates for cost-sensitive learning on manifolds with approximate
nearest neighbours | We study the approximate nearest neighbour method for cost-sensitive
classification on low-dimensional manifolds embedded within a high-dimensional
feature space. We determine the minimax learning rates for distributions on a
smooth manifold, in a cost-sensitive setting. This generalises a classic result
of Audibert and Tsybakov. Building upon recent work of Chaudhuri and Dasgupta
we prove that these minimax rates are attained by the approximate nearest
neighbour algorithm, where neighbours are computed in a randomly projected
low-dimensional space. In addition, we give a bound on the number of dimensions
required for the projection which depends solely upon the reach and dimension
of the manifold, combined with the regularity of the marginal.
|
Diversity and degrees of freedom in regression ensembles | Ensemble methods are a cornerstone of modern machine learning. The
performance of an ensemble depends crucially upon the level of diversity
between its constituent learners. This paper establishes a connection between
diversity and degrees of freedom (i.e. the capacity of the model), showing that
diversity may be viewed as a form of inverse regularisation. This is achieved
by focusing on a previously published algorithm Negative Correlation Learning
(NCL), in which model diversity is explicitly encouraged through a diversity
penalty term in the loss function. We provide an exact formula for the
effective degrees of freedom in an NCL ensemble with fixed basis functions,
showing that it is a continuous, convex and monotonically increasing function
of the diversity parameter. We demonstrate a connection to Tikhonov
regularisation and show that, with an appropriately chosen diversity parameter,
an NCL ensemble can always outperform the unregularised ensemble in the
presence of noise. We demonstrate the practical utility of our approach by
deriving a method to efficiently tune the diversity parameter. Finally, we use
a Monte-Carlo estimator to extend the connection between diversity and degrees
of freedom to ensembles of deep neural networks.
|
The K-Nearest Neighbour UCB algorithm for multi-armed bandits with
covariates | In this paper we propose and explore the k-Nearest Neighbour UCB algorithm
for multi-armed bandits with covariates. We focus on a setting where the
covariates are supported on a metric space of low intrinsic dimension, such as
a manifold embedded within a high dimensional ambient feature space. The
algorithm is conceptually simple and straightforward to implement. The
k-Nearest Neighbour UCB algorithm does not require prior knowledge of the
either the intrinsic dimension of the marginal distribution or the time
horizon. We prove a regret bound for the k-Nearest Neighbour UCB algorithm
which is minimax optimal up to logarithmic factors. In particular, the
algorithm automatically takes advantage of both low intrinsic dimensionality of
the marginal distribution over the covariates and low noise in the data,
expressed as a margin condition. In addition, focusing on the case of bounded
rewards, we give corresponding regret bounds for the k-Nearest Neighbour KL-UCB
algorithm, which is an analogue of the KL-UCB algorithm adapted to the setting
of multi-armed bandits with covariates. Finally, we present empirical results
which demonstrate the ability of both the k-Nearest Neighbour UCB and k-Nearest
Neighbour KL-UCB to take advantage of situations where the data is supported on
an unknown sub-manifold of a high-dimensional feature space.
|
Satellite imagery analysis for operational damage assessment in
Emergency situations | When major disaster occurs the questions are raised how to estimate the
damage in time to support the decision making process and relief efforts by
local authorities or humanitarian teams. In this paper we consider the use of
Machine Learning and Computer Vision on remote sensing imagery to improve time
efficiency of assessment of damaged buildings in disaster affected area. We
propose a general workflow that can be useful in various disaster management
applications, and demonstrate the use of the proposed workflow for the
assessment of the damage caused by the wildfires in California in 2017.
|
Calcium Removal From Cardiac CT Images Using Deep Convolutional Neural
Network | Coronary calcium causes beam hardening and blooming artifacts on cardiac
computed tomography angiography (CTA) images, which lead to overestimation of
lumen stenosis and reduction of diagnostic specificity. To properly remove
coronary calcification and restore arterial lumen precisely, we propose a
machine learning-based method with a multi-step inpainting process. We
developed a new network configuration, Dense-Unet, to achieve optimal
performance with low computational cost. Results after the calcium removal
process were validated by comparing with gold-standard X-ray angiography. Our
results demonstrated that removing coronary calcification from images with the
proposed approach was feasible, and may potentially improve the diagnostic
accuracy of CTA.
|
Deep Defense: Training DNNs with Improved Adversarial Robustness | Despite the efficacy on a variety of computer vision tasks, deep neural
networks (DNNs) are vulnerable to adversarial attacks, limiting their
applications in security-critical systems. Recent works have shown the
possibility of generating imperceptibly perturbed image inputs (a.k.a.,
adversarial examples) to fool well-trained DNN classifiers into making
arbitrary predictions. To address this problem, we propose a training recipe
named "deep defense". Our core idea is to integrate an adversarial
perturbation-based regularizer into the classification objective, such that the
obtained models learn to resist potential attacks, directly and precisely. The
whole optimization problem is solved just like training a recursive network.
Experimental results demonstrate that our method outperforms training with
adversarial/Parseval regularizations by large margins on various datasets
(including MNIST, CIFAR-10 and ImageNet) and different DNN architectures. Code
and models for reproducing our results are available at
https://github.com/ZiangYan/deepdefense.pytorch
|
Tractable and Scalable Schatten Quasi-Norm Approximations for Rank
Minimization | The Schatten quasi-norm was introduced to bridge the gap between the trace
norm and rank function. However, existing algorithms are too slow or even
impractical for large-scale problems. Motivated by the equivalence relation
between the trace norm and its bilinear spectral penalty, we define two
tractable Schatten norms, i.e.\ the bi-trace and tri-trace norms, and prove
that they are in essence the Schatten-$1/2$ and $1/3$ quasi-norms,
respectively. By applying the two defined Schatten quasi-norms to various rank
minimization problems such as MC and RPCA, we only need to solve much smaller
factor matrices. We design two efficient linearized alternating minimization
algorithms to solve our problems and establish that each bounded sequence
generated by our algorithms converges to a critical point. We also provide the
restricted strong convexity (RSC) based and MC error bounds for our algorithms.
Our experimental results verified both the efficiency and effectiveness of our
algorithms compared with the state-of-the-art methods.
|
Knowledge Transfer with Jacobian Matching | Classical distillation methods transfer representations from a "teacher"
neural network to a "student" network by matching their output activations.
Recent methods also match the Jacobians, or the gradient of output activations
with the input. However, this involves making some ad hoc decisions, in
particular, the choice of the loss function.
In this paper, we first establish an equivalence between Jacobian matching
and distillation with input noise, from which we derive appropriate loss
functions for Jacobian matching. We then rely on this analysis to apply
Jacobian matching to transfer learning by establishing equivalence of a recent
transfer learning procedure to distillation.
We then show experimentally on standard image datasets that Jacobian-based
penalties improve distillation, robustness to noisy inputs, and transfer
learning.
|
Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal
Modeling | Advances in the field of inverse reinforcement learning (IRL) have led to
sophisticated inference frameworks that relax the original modeling assumption
of observing an agent behavior that reflects only a single intention. Instead
of learning a global behavioral model, recent IRL methods divide the
demonstration data into parts, to account for the fact that different
trajectories may correspond to different intentions, e.g., because they were
generated by different domain experts. In this work, we go one step further:
using the intuitive concept of subgoals, we build upon the premise that even a
single trajectory can be explained more efficiently locally within a certain
context than globally, enabling a more compact representation of the observed
behavior. Based on this assumption, we build an implicit intentional model of
the agent's goals to forecast its behavior in unobserved situations. The result
is an integrated Bayesian prediction framework that significantly outperforms
existing IRL solutions and provides smooth policy estimates consistent with the
expert's plan. Most notably, our framework naturally handles situations where
the intentions of the agent change over time and classical IRL algorithms fail.
In addition, due to its probabilistic nature, the model can be
straightforwardly applied in active learning scenarios to guide the
demonstration process of the expert.
|
Inferring Missing Categorical Information in Noisy and Sparse Web Markup | Embedded markup of Web pages has seen widespread adoption throughout the past
years driven by standards such as RDFa and Microdata and initiatives such as
schema.org, where recent studies show an adoption by 39% of all Web pages
already in 2016. While this constitutes an important information source for
tasks such as Web search, Web page classification or knowledge graph
augmentation, individual markup nodes are usually sparsely described and often
lack essential information. For instance, from 26 million nodes describing
events within the Common Crawl in 2016, 59% of nodes provide less than six
statements and only 257,000 nodes (0.96%) are typed with more specific event
subtypes. Nevertheless, given the scale and diversity of Web markup data, nodes
that provide missing information can be obtained from the Web in large
quantities, in particular for categorical properties. Such data constitutes
potential training data for inferring missing information to significantly
augment sparsely described nodes. In this work, we introduce a supervised
approach for inferring missing categorical properties in Web markup. Our
experiments, conducted on properties of events and movies, show a performance
of 79% and 83% F1 score correspondingly, significantly outperforming existing
baselines.
|
The Power Mean Laplacian for Multilayer Graph Clustering | Multilayer graphs encode different kind of interactions between the same set
of entities. When one wants to cluster such a multilayer graph, the natural
question arises how one should merge the information different layers. We
introduce in this paper a one-parameter family of matrix power means for
merging the Laplacians from different layers and analyze it in expectation in
the stochastic block model. We show that this family allows to recover ground
truth clusters under different settings and verify this in real world data.
While computing the matrix power mean can be very expensive for large graphs,
we introduce a numerical scheme to efficiently compute its eigenvectors for the
case of large sparse graphs.
|
Natural data structure extracted from neighborhood-similarity graphs | 'Big' high-dimensional data are commonly analyzed in low-dimensions, after
performing a dimensionality-reduction step that inherently distorts the data
structure. For the same purpose, clustering methods are also often used. These
methods also introduce a bias, either by starting from the assumption of a
particular geometric form of the clusters, or by using iterative schemes to
enhance cluster contours, with uncontrollable consequences. The goal of data
analysis should, however, be to encode and detect structural data features at
all scales and densities simultaneously, without assuming a parametric form of
data point distances, or modifying them. We propose a novel approach that
directly encodes data point neighborhood similarities as a sparse graph. Our
non-iterative framework permits a transparent interpretation of data, without
altering the original data dimension and metric. Several natural and synthetic
data applications demonstrate the efficacy of our novel approach.
|
Understand Functionality and Dimensionality of Vector Embeddings: the
Distributional Hypothesis, the Pairwise Inner Product Loss and Its
Bias-Variance Trade-off | Vector embedding is a foundational building block of many deep learning
models, especially in natural language processing. In this paper, we present a
theoretical framework for understanding the effect of dimensionality on vector
embeddings. We observe that the distributional hypothesis, a governing
principle of statistical semantics, requires a natural unitary-invariance for
vector embeddings. Motivated by the unitary-invariance observation, we propose
the Pairwise Inner Product (PIP) loss, a unitary-invariant metric on the
similarity between two embeddings. We demonstrate that the PIP loss captures
the difference in functionality between embeddings, and that the PIP loss is
tightly connect with two basic properties of vector embeddings, namely
similarity and compositionality. By formulating the embedding training process
as matrix factorization with noise, we reveal a fundamental bias-variance
trade-off between the signal spectrum and noise power in the dimensionality
selection process. This bias-variance trade-off sheds light on many empirical
observations which have not been thoroughly explained, for example the
existence of an optimal dimensionality. Moreover, we discover two new results
about vector embeddings, namely their robustness against over-parametrization
and their forward stability. The bias-variance trade-off of the PIP loss
explicitly answers the fundamental open problem of dimensionality selection for
vector embeddings.
|
Online Feature Ranking for Intrusion Detection Systems | Many current approaches to the design of intrusion detection systems apply
feature selection in a static, non-adaptive fashion. These methods often
neglect the dynamic nature of network data which requires to use adaptive
feature selection techniques. In this paper, we present a simple technique
based on incremental learning of support vector machines in order to rank the
features in real time within a streaming model for network data. Some
illustrative numerical experiments with two popular benchmark datasets show
that our approach allows to adapt to the changes in normal network behaviour
and novel attack patterns which have not been experienced before.
|
Semi-Supervised Online Structure Learning for Composite Event
Recognition | Online structure learning approaches, such as those stemming from Statistical
Relational Learning, enable the discovery of complex relations in noisy data
streams. However, these methods assume the existence of fully-labelled training
data, which is unrealistic for most real-world applications. We present a novel
approach for completing the supervision of a semi-supervised structure learning
task. We incorporate graph-cut minimisation, a technique that derives labels
for unlabelled data, based on their distance to their labelled counterparts. In
order to adapt graph-cut minimisation to first order logic, we employ a
suitable structural distance for measuring the distance between sets of logical
atoms. The labelling process is achieved online (single-pass) by means of a
caching mechanism and the Hoeffding bound, a statistical tool to approximate
globally-optimal decisions from locally-optimal ones. We evaluate our approach
on the task of composite event recognition by using a benchmark dataset for
human activity recognition, as well as a real dataset for maritime monitoring.
The evaluation suggests that our approach can effectively complete the missing
labels and eventually, improve the accuracy of the underlying structure
learning system.
|
Hierarchical Imitation and Reinforcement Learning | We study how to effectively leverage expert feedback to learn sequential
decision-making policies. We focus on problems with sparse rewards and long
time horizons, which typically pose significant challenges in reinforcement
learning. We propose an algorithmic framework, called hierarchical guidance,
that leverages the hierarchical structure of the underlying problem to
integrate different modes of expert interaction. Our framework can incorporate
different combinations of imitation learning (IL) and reinforcement learning
(RL) at different levels, leading to dramatic reductions in both expert effort
and cost of exploration. Using long-horizon benchmarks, including Montezuma's
Revenge, we demonstrate that our approach can learn significantly faster than
hierarchical RL, and be significantly more label-efficient than standard IL. We
also theoretically analyze labeling cost for certain instantiations of our
framework.
|
On Oracle-Efficient PAC RL with Rich Observations | We study the computational tractability of PAC reinforcement learning with
rich observations. We present new provably sample-efficient algorithms for
environments with deterministic hidden state dynamics and stochastic rich
observations. These methods operate in an oracle model of computation --
accessing policy and value function classes exclusively through standard
optimization primitives -- and therefore represent computationally efficient
alternatives to prior algorithms that require enumeration. With stochastic
hidden state dynamics, we prove that the only known sample-efficient algorithm,
OLIVE, cannot be implemented in the oracle model. We also present several
examples that illustrate fundamental challenges of tractable PAC reinforcement
learning in such general settings.
|
Static and Dynamic Robust PCA and Matrix Completion: A Review | Principal Components Analysis (PCA) is one of the most widely used dimension
reduction techniques. Robust PCA (RPCA) refers to the problem of PCA when the
data may be corrupted by outliers. Recent work by Cand{\`e}s, Wright, Li, and
Ma defined RPCA as a problem of decomposing a given data matrix into the sum of
a low-rank matrix (true data) and a sparse matrix (outliers). The column space
of the low-rank matrix then gives the PCA solution. This simple definition has
lead to a large amount of interesting new work on provably correct, fast, and
practical solutions to RPCA. More recently, the dynamic (time-varying) version
of the RPCA problem has been studied and a series of provably correct, fast,
and memory efficient tracking solutions have been proposed. Dynamic RPCA (or
robust subspace tracking) is the problem of tracking data lying in a (slowly)
changing subspace while being robust to sparse outliers. This article provides
an exhaustive review of the last decade of literature on RPCA and its dynamic
counterpart (robust subspace tracking), along with describing their theoretical
guarantees, discussing the pros and cons of various approaches, and providing
empirical comparisons of performance and speed.
A brief overview of the (low-rank) matrix completion literature is also
provided (the focus is on works not discussed in other recent reviews). This
refers to the problem of completing a low-rank matrix when only a subset of its
entries are observed. It can be interpreted as a simpler special case of RPCA
in which the indices of the outlier corrupted entries are known.
|
Semi-parametric Topological Memory for Navigation | We introduce a new memory architecture for navigation in previously unseen
environments, inspired by landmark-based navigation in animals. The proposed
semi-parametric topological memory (SPTM) consists of a (non-parametric) graph
with nodes corresponding to locations in the environment and a (parametric)
deep network capable of retrieving nodes from the graph based on observations.
The graph stores no metric information, only connectivity of locations
corresponding to the nodes. We use SPTM as a planning module in a navigation
system. Given only 5 minutes of footage of a previously unseen maze, an
SPTM-based navigation agent can build a topological map of the environment and
use it to confidently navigate towards goals. The average success rate of the
SPTM agent in goal-directed navigation across test environments is higher than
the best-performing baseline by a factor of three. A video of the agent is
available at https://youtu.be/vRF7f4lhswo
|
Evolutionary Generative Adversarial Networks | Generative adversarial networks (GAN) have been effective for learning
generative models for real-world data. However, existing GANs (GAN and its
variants) tend to suffer from training problems such as instability and mode
collapse. In this paper, we propose a novel GAN framework called evolutionary
generative adversarial networks (E-GAN) for stable GAN training and improved
generative performance. Unlike existing GANs, which employ a pre-defined
adversarial objective function alternately training a generator and a
discriminator, we utilize different adversarial training objectives as mutation
operations and evolve a population of generators to adapt to the environment
(i.e., the discriminator). We also utilize an evaluation mechanism to measure
the quality and diversity of generated samples, such that only well-performing
generator(s) are preserved and used for further training. In this way, E-GAN
overcomes the limitations of an individual adversarial training objective and
always preserves the best offspring, contributing to progress in and the
success of GANs. Experiments on several datasets demonstrate that E-GAN
achieves convincing generative performance and reduces the training problems
inherent in existing GANs.
|
Meta-Learning for Semi-Supervised Few-Shot Classification | In few-shot classification, we are interested in learning algorithms that
train a classifier from only a handful of labeled examples. Recent progress in
few-shot classification has featured meta-learning, in which a parameterized
model for a learning algorithm is defined and trained on episodes representing
different classification problems, each with a small labeled training set and
its corresponding test set. In this work, we advance this few-shot
classification paradigm towards a scenario where unlabeled examples are also
available within each episode. We consider two situations: one where all
unlabeled examples are assumed to belong to the same set of classes as the
labeled examples of the episode, as well as the more challenging situation
where examples from other distractor classes are also provided. To address this
paradigm, we propose novel extensions of Prototypical Networks (Snell et al.,
2017) that are augmented with the ability to use unlabeled examples when
producing prototypes. These models are trained in an end-to-end way on
episodes, to learn to leverage the unlabeled examples successfully. We evaluate
these methods on versions of the Omniglot and miniImageNet benchmarks, adapted
to this new framework augmented with unlabeled examples. We also propose a new
split of ImageNet, consisting of a large set of classes, with a hierarchical
structure. Our experiments confirm that our Prototypical Networks can learn to
improve their predictions due to unlabeled examples, much like a
semi-supervised algorithm would.
|
Random perturbation and matrix sparsification and completion | We discuss general perturbation inequalities when the perturbation is random.
As applications, we obtain several new results concerning two important
problems: matrix sparsification and matrix completion.
|
Autostacker: A Compositional Evolutionary Learning System | We introduce an automatic machine learning (AutoML) modeling architecture
called Autostacker, which combines an innovative hierarchical stacking
architecture and an Evolutionary Algorithm (EA) to perform efficient parameter
search. Neither prior domain knowledge about the data nor feature preprocessing
is needed. Using EA, Autostacker quickly evolves candidate pipelines with high
predictive accuracy. These pipelines can be used as is or as a starting point
for human experts to build on. Autostacker finds innovative combinations and
structures of machine learning models, rather than selecting a single model and
optimizing its hyperparameters. Compared with other AutoML systems on fifteen
datasets, Autostacker achieves state-of-art or competitive performance both in
terms of test accuracy and time cost.
|
Raw Multi-Channel Audio Source Separation using Multi-Resolution
Convolutional Auto-Encoders | Supervised multi-channel audio source separation requires extracting useful
spectral, temporal, and spatial features from the mixed signals. The success of
many existing systems is therefore largely dependent on the choice of features
used for training. In this work, we introduce a novel multi-channel,
multi-resolution convolutional auto-encoder neural network that works on raw
time-domain signals to determine appropriate multi-resolution features for
separating the singing-voice from stereo music. Our experimental results show
that the proposed method can achieve multi-channel audio source separation
without the need for hand-crafted features or any pre- or post-processing.
|
Reinforcement Learning to Rank in E-Commerce Search Engine:
Formalization, Analysis, and Application | In e-commerce platforms such as Amazon and TaoBao, ranking items in a search
session is a typical multi-step decision-making problem. Learning to rank (LTR)
methods have been widely applied to ranking problems. However, such methods
often consider different ranking steps in a session to be independent, which
conversely may be highly correlated to each other. For better utilizing the
correlation between different ranking steps, in this paper, we propose to use
reinforcement learning (RL) to learn an optimal ranking policy which maximizes
the expected accumulative rewards in a search session. Firstly, we formally
define the concept of search session Markov decision process (SSMDP) to
formulate the multi-step ranking problem. Secondly, we analyze the property of
SSMDP and theoretically prove the necessity of maximizing accumulative rewards.
Lastly, we propose a novel policy gradient algorithm for learning an optimal
ranking policy, which is able to deal with the problem of high reward variance
and unbalanced reward distribution of an SSMDP. Experiments are conducted in
simulation and TaoBao search engine. The results demonstrate that our algorithm
performs much better than online LTR methods, with more than 40% and 30% growth
of total transaction amount in the simulation and the real application,
respectively.
|
Clinically Meaningful Comparisons Over Time: An Approach to Measuring
Patient Similarity based on Subsequence Alignment | Longitudinal patient data has the potential to improve clinical risk
stratification models for disease. However, chronic diseases that progress
slowly over time are often heterogeneous in their clinical presentation.
Patients may progress through disease stages at varying rates. This leads to
pathophysiological misalignment over time, making it difficult to consistently
compare patients in a clinically meaningful way. Furthermore, patients present
clinically for the first time at different stages of disease. This eliminates
the possibility of simply aligning patients based on their initial
presentation. Finally, patient data may be sampled at different rates due to
differences in schedules or missed visits. To address these challenges, we
propose a robust measure of patient similarity based on subsequence alignment.
Compared to global alignment techniques that do not account for
pathophysiological misalignment, focusing on the most relevant subsequences
allows for an accurate measure of similarity between patients. We demonstrate
the utility of our approach in settings where longitudinal data, while useful,
are limited and lack a clear temporal alignment for comparison. Applied to the
task of stratifying patients for risk of progression to probable Alzheimer's
Disease, our approach outperforms models that use only snapshot data (AUROC of
0.839 vs. 0.812) and models that use global alignment techniques (AUROC of
0.822). Our results support the hypothesis that patients' trajectories are
useful for quantifying inter-patient similarities and that using subsequence
matching and can help account for heterogeneity and misalignment in
longitudinal data.
|
Convolutional Geometric Matrix Completion | Geometric matrix completion (GMC) has been proposed for recommendation by
integrating the relationship (link) graphs among users/items into matrix
completion (MC). Traditional GMC methods typically adopt graph regularization
to impose smoothness priors for MC. Recently, geometric deep learning on graphs
(GDLG) is proposed to solve the GMC problem, showing better performance than
existing GMC methods including traditional graph regularization based methods.
To the best of our knowledge, there exists only one GDLG method for GMC, which
is called RMGCNN. RMGCNN combines graph convolutional network (GCN) and
recurrent neural network (RNN) together for GMC. In the original work of
RMGCNN, RMGCNN demonstrates better performance than pure GCN-based method. In
this paper, we propose a new GMC method, called convolutional geometric matrix
completion (CGMC), for recommendation with graphs among users/items. CGMC is a
pure GCN-based method with a newly designed graph convolutional network.
Experimental results on real datasets show that CGMC can outperform other
state-of-the-art methods including RMGCNN in terms of both accuracy and speed.
|
Unsupervised Learning of Goal Spaces for Intrinsically Motivated Goal
Exploration | Intrinsically motivated goal exploration algorithms enable machines to
discover repertoires of policies that produce a diversity of effects in complex
environments. These exploration algorithms have been shown to allow real world
robots to acquire skills such as tool use in high-dimensional continuous state
and action spaces. However, they have so far assumed that self-generated goals
are sampled in a specifically engineered feature space, limiting their
autonomy. In this work, we propose to use deep representation learning
algorithms to learn an adequate goal space. This is a developmental 2-stage
approach: first, in a perceptual learning stage, deep learning algorithms use
passive raw sensor observations of world changes to learn a corresponding
latent space; then goal exploration happens in a second stage by sampling goals
in this latent space. We present experiments where a simulated robot arm
interacts with an object, and we show that exploration algorithms using such
learned representations can match the performance obtained using engineered
representations.
|
Detecting non-causal artifacts in multivariate linear regression models | We consider linear models where $d$ potential causes $X_1,...,X_d$ are
correlated with one target quantity $Y$ and propose a method to infer whether
the association is causal or whether it is an artifact caused by overfitting or
hidden common causes. We employ the idea that in the former case the vector of
regression coefficients has 'generic' orientation relative to the covariance
matrix $\Sigma_{XX}$ of $X$. Using an ICA based model for confounding, we show
that both confounding and overfitting yield regression vectors that concentrate
mainly in the space of low eigenvalues of $\Sigma_{XX}$.
|
NetGAN: Generating Graphs via Random Walks | We propose NetGAN - the first implicit generative model for graphs able to
mimic real-world networks. We pose the problem of graph generation as learning
the distribution of biased random walks over the input graph. The proposed
model is based on a stochastic neural network that generates discrete output
samples and is trained using the Wasserstein GAN objective. NetGAN is able to
produce graphs that exhibit well-known network patterns without explicitly
specifying them in the model definition. At the same time, our model exhibits
strong generalization properties, as highlighted by its competitive link
prediction performance, despite not being trained specifically for this task.
Being the first approach to combine both of these desirable properties, NetGAN
opens exciting avenues for further research.
|
Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with
Category Shift | Unsupervised domain adaptation (UDA) conventionally assumes labeled source
samples coming from a single underlying source distribution. Whereas in
practical scenario, labeled data are typically collected from diverse sources.
The multiple sources are different not only from the target but also from each
other, thus, domain adaptater should not be modeled in the same way. Moreover,
those sources may not completely share their categories, which further brings a
new transfer challenge called category shift. In this paper, we propose a deep
cocktail network (DCTN) to battle the domain and category shifts among multiple
sources. Motivated by the theoretical results in \cite{mansour2009domain}, the
target distribution can be represented as the weighted combination of source
distributions, and, the multi-source unsupervised domain adaptation via DCTN is
then performed as two alternating steps: i) It deploys multi-way adversarial
learning to minimize the discrepancy between the target and each of the
multiple source domains, which also obtains the source-specific perplexity
scores to denote the possibilities that a target sample belongs to different
source domains. ii) The multi-source category classifiers are integrated with
the perplexity scores to classify target sample, and the pseudo-labeled target
samples together with source samples are utilized to update the multi-source
category classifier and the feature extractor. We evaluate DCTN in three domain
adaptation benchmarks, which clearly demonstrate the superiority of our
framework.
|
A multi-instance deep neural network classifier: application to Higgs
boson CP measurement | We investigate properties of a classifier applied to the measurements of the
CP state of the Higgs boson in $H\rightarrow\tau\tau$ decays. The problem is
framed as binary classifier applied to individual instances. Then the prior
knowledge that the instances belong to the same class is used to define the
multi-instance classifier. Its final score is calculated as multiplication of
single instance scores for a given series of instances. In the paper we discuss
properties of such classifier, notably its dependence on the number of
instances in the series. This classifier exhibits very strong random dependence
on the number of epochs used for training and requires careful tuning of the
classification threshold. We derive formula for this optimal threshold.
|
Gradient-based Sampling: An Adaptive Importance Sampling for
Least-squares | In modern data analysis, random sampling is an efficient and widely-used
strategy to overcome the computational difficulties brought by large sample
size. In previous studies, researchers conducted random sampling which is
according to the input data but independent on the response variable, however
the response variable may also be informative for sampling. In this paper we
propose an adaptive sampling called the gradient-based sampling which is
dependent on both the input data and the output for fast solving of
least-square (LS) problems. We draw the data points by random sampling from the
full data according to their gradient values. This sampling is computationally
saving, since the running time of computing the sampling probabilities is
reduced to O(nd) where n is the full sample size and d is the dimension of the
input. Theoretically, we establish an error bound analysis of the general
importance sampling with respect to LS solution from full data. The result
establishes an improved performance of the use of our gradient- based sampling.
Synthetic and real data sets are used to empirically argue that the
gradient-based sampling has an obvious advantage over existing sampling methods
from two aspects of statistical efficiency and computational saving.
|
A more globally accurate dimensionality reduction method using triplets | We first show that the commonly used dimensionality reduction (DR) methods
such as t-SNE and LargeVis poorly capture the global structure of the data in
the low dimensional embedding. We show this via a number of tests for the DR
methods that can be easily applied by any practitioner to the dataset at hand.
Surprisingly enough, t-SNE performs the best w.r.t. the commonly used measures
that reward the local neighborhood accuracy such as precision-recall while
having the worst performance in our tests for global structure. We then
contrast the performance of these two DR method against our new method called
TriMap. The main idea behind TriMap is to capture higher orders of structure
with triplet information (instead of pairwise information used by t-SNE and
LargeVis), and to minimize a robust loss function for satisfying the chosen
triplets. We provide compelling experimental evidence on large natural datasets
for the clear advantage of the TriMap DR results. As LargeVis, TriMap scales
linearly with the number of data points.
|
Probabilistic design of a molybdenum-base alloy using a neural network | An artificial intelligence tool is exploited to discover and characterize a
new molybdenum-base alloy that is the most likely to simultaneously satisfy
targets of cost, phase stability, precipitate content, yield stress, and
hardness. Experimental testing demonstrates that the proposed alloy fulfils the
computational predictions, and furthermore the physical properties exceed those
of other commercially available Mo-base alloys for forging-die applications.
|
Essentially No Barriers in Neural Network Energy Landscape | Training neural networks involves finding minima of a high-dimensional
non-convex loss function. Knowledge of the structure of this energy landscape
is sparse. Relaxing from linear interpolations, we construct continuous paths
between minima of recent neural network architectures on CIFAR10 and CIFAR100.
Surprisingly, the paths are essentially flat in both the training and test
landscapes. This implies that neural networks have enough capacity for
structural changes, or that these changes are small between minima. Also, each
minimum has at least one vanishing Hessian eigenvalue in addition to those
resulting from trivial invariance.
|
Deep factorization for speech signal | Various informative factors mixed in speech signals, leading to great
difficulty when decoding any of the factors. An intuitive idea is to factorize
each speech frame into individual informative factors, though it turns out to
be highly difficult. Recently, we found that speaker traits, which were assumed
to be long-term distributional properties, are actually short-time patterns,
and can be learned by a carefully designed deep neural network (DNN). This
discovery motivated a cascade deep factorization (CDF) framework that will be
presented in this paper. The proposed framework infers speech factors in a
sequential way, where factors previously inferred are used as conditional
variables when inferring other factors. We will show that this approach can
effectively factorize speech signals, and using these factors, the original
speech spectrum can be recovered with a high accuracy. This factorization and
reconstruction approach provides potential values for many speech processing
tasks, e.g., speaker recognition and emotion recognition, as will be
demonstrated in the paper.
|
Impact of Biases in Big Data | The underlying paradigm of big data-driven machine learning reflects the
desire of deriving better conclusions from simply analyzing more data, without
the necessity of looking at theory and models. Is having simply more data
always helpful? In 1936, The Literary Digest collected 2.3M filled in
questionnaires to predict the outcome of that year's US presidential election.
The outcome of this big data prediction proved to be entirely wrong, whereas
George Gallup only needed 3K handpicked people to make an accurate prediction.
Generally, biases occur in machine learning whenever the distributions of
training set and test set are different. In this work, we provide a review of
different sorts of biases in (big) data sets in machine learning. We provide
definitions and discussions of the most commonly appearing biases in machine
learning: class imbalance and covariate shift. We also show how these biases
can be quantified and corrected. This work is an introductory text for both
researchers and practitioners to become more aware of this topic and thus to
derive more reliable models for their learning problems.
|
Understanding the Loss Surface of Neural Networks for Binary
Classification | It is widely conjectured that the reason that training algorithms for neural
networks are successful because all local minima lead to similar performance,
for example, see (LeCun et al., 2015, Choromanska et al., 2015, Dauphin et al.,
2014). Performance is typically measured in terms of two metrics: training
performance and generalization performance. Here we focus on the training
performance of single-layered neural networks for binary classification, and
provide conditions under which the training error is zero at all local minima
of a smooth hinge loss function. Our conditions are roughly in the following
form: the neurons have to be strictly convex and the surrogate loss function
should be a smooth version of hinge loss. We also provide counterexamples to
show that when the loss function is replaced with quadratic loss or logistic
loss, the result may not hold.
|
Deep Learning for Signal Authentication and Security in Massive Internet
of Things Systems | Secure signal authentication is arguably one of the most challenging problems
in the Internet of Things (IoT) environment, due to the large-scale nature of
the system and its susceptibility to man-in-the-middle and eavesdropping
attacks. In this paper, a novel deep learning method is proposed for dynamic
authentication of IoT signals to detect cyber attacks. The proposed learning
framework, based on a long short-term memory (LSTM) structure, enables the IoT
devices (IoTDs) to extract a set of stochastic features from their generated
signal and dynamically watermark these features into the signal. This method
enables the cloud, which collects signals from the IoT devices, to effectively
authenticate the reliability of the signals. Moreover, in massive IoT
scenarios, since the cloud cannot authenticate all the IoTDs simultaneously due
to computational limitations, a game-theoretic framework is proposed to improve
the cloud's decision making process by predicting vulnerable IoTDs. The
mixed-strategy Nash equilibrium (MSNE) for this game is derived and the
uniqueness of the expected utility at the equilibrium is proven. In the massive
IoT system, due to a large set of available actions for the cloud, it is shown
that analytically deriving the MSNE is challenging and, thus, a learning
algorithm proposed that converges to the MSNE. Moreover, in order to cope with
the incomplete information case in which the cloud cannot access the state of
the unauthenticated IoTDs, a deep reinforcement learning algorithm is proposed
to dynamically predict the state of unauthenticated IoTDs and allow the cloud
to decide on which IoTDs to authenticate. Simulation results show that, with an
attack detection delay of under 1 second the messages can be transmitted from
IoT devices with an almost 100% reliability.
|
Distributed Prioritized Experience Replay | We propose a distributed architecture for deep reinforcement learning at
scale, that enables agents to learn effectively from orders of magnitude more
data than previously possible. The algorithm decouples acting from learning:
the actors interact with their own instances of the environment by selecting
actions according to a shared neural network, and accumulate the resulting
experience in a shared experience replay memory; the learner replays samples of
experience and updates the neural network. The architecture relies on
prioritized experience replay to focus only on the most significant data
generated by the actors. Our architecture substantially improves the state of
the art on the Arcade Learning Environment, achieving better final performance
in a fraction of the wall-clock training time.
|
Not All Samples Are Created Equal: Deep Learning with Importance
Sampling | Deep neural network training spends most of the computation on examples that
are properly handled, and could be ignored. We propose to mitigate this
phenomenon with a principled importance sampling scheme that focuses
computation on "informative" examples, and reduces the variance of the
stochastic gradients during training. Our contribution is twofold: first, we
derive a tractable upper bound to the per-sample gradient norm, and second we
derive an estimator of the variance reduction achieved with importance
sampling, which enables us to switch it on when it will result in an actual
speedup. The resulting scheme can be used by changing a few lines of code in a
standard SGD procedure, and we demonstrate experimentally, on image
classification, CNN fine-tuning, and RNN training, that for a fixed wall-clock
time budget, it provides a reduction of the train losses of up to an order of
magnitude and a relative improvement of test errors between 5% and 17%.
|
Active model learning and diverse action sampling for task and motion
planning | The objective of this work is to augment the basic abilities of a robot by
learning to use new sensorimotor primitives to enable the solution of complex
long-horizon problems. Solving long-horizon problems in complex domains
requires flexible generative planning that can combine primitive abilities in
novel combinations to solve problems as they arise in the world. In order to
plan to combine primitive actions, we must have models of the preconditions and
effects of those actions: under what circumstances will executing this
primitive achieve some particular effect in the world?
We use, and develop novel improvements on, state-of-the-art methods for
active learning and sampling. We use Gaussian process methods for learning the
conditions of operator effectiveness from small numbers of expensive training
examples collected by experimentation on a robot. We develop adaptive sampling
methods for generating diverse elements of continuous sets (such as robot
configurations and object poses) during planning for solving a new task, so
that planning is as efficient as possible. We demonstrate these methods in an
integrated system, combining newly learned models with an efficient
continuous-space robot task and motion planner to learn to solve long horizon
problems more efficiently than was previously possible.
|
Label Sanitization against Label Flipping Poisoning Attacks | Many machine learning systems rely on data collected in the wild from
untrusted sources, exposing the learning algorithms to data poisoning.
Attackers can inject malicious data in the training dataset to subvert the
learning process, compromising the performance of the algorithm producing
errors in a targeted or an indiscriminate way. Label flipping attacks are a
special case of data poisoning, where the attacker can control the labels
assigned to a fraction of the training points. Even if the capabilities of the
attacker are constrained, these attacks have been shown to be effective to
significantly degrade the performance of the system. In this paper we propose
an efficient algorithm to perform optimal label flipping poisoning attacks and
a mechanism to detect and relabel suspicious data points, mitigating the effect
of such poisoning attacks.
|
An Overview of Robust Subspace Recovery | This paper will serve as an introduction to the body of work on robust
subspace recovery. Robust subspace recovery involves finding an underlying
low-dimensional subspace in a dataset that is possibly corrupted with outliers.
While this problem is easy to state, it has been difficult to develop optimal
algorithms due to its underlying nonconvexity. This work emphasizes advantages
and disadvantages of proposed approaches and unsolved problems in the area.
|
PRESISTANT: Learning based assistant for data pre-processing | Data pre-processing is one of the most time consuming and relevant steps in a
data analysis process (e.g., classification task). A given data pre-processing
operator (e.g., transformation) can have positive, negative or zero impact on
the final result of the analysis. Expert users have the required knowledge to
find the right pre-processing operators. However, when it comes to non-experts,
they are overwhelmed by the amount of pre-processing operators and it is
challenging for them to find operators that would positively impact their
analysis (e.g., increase the predictive accuracy of a classifier). Existing
solutions either assume that users have expert knowledge, or they recommend
pre-processing operators that are only "syntactically" applicable to a dataset,
without taking into account their impact on the final analysis. In this work,
we aim at providing assistance to non-expert users by recommending data
pre-processing operators that are ranked according to their impact on the final
analysis. We developed a tool PRESISTANT, that uses Random Forests to learn the
impact of pre-processing operators on the performance (e.g., predictive
accuracy) of 5 different classification algorithms, such as J48, Naive Bayes,
PART, Logistic Regression, and Nearest Neighbor. Extensive evaluations on the
recommendations provided by our tool, show that PRESISTANT can effectively help
non-experts in order to achieve improved results in their analytical tasks.
|
Building a Telescope to Look Into High-Dimensional Image Spaces | An image pattern can be represented by a probability distribution whose
density is concentrated on different low-dimensional subspaces in the
high-dimensional image space. Such probability densities have an astronomical
number of local modes corresponding to typical pattern appearances. Related
groups of modes can join to form macroscopic image basins that represent
pattern concepts. Recent works use neural networks that capture high-order
image statistics to learn Gibbs models capable of synthesizing realistic images
of many patterns. However, characterizing a learned probability density to
uncover the Hopfield memories of the model, encoded by the structure of the
local modes, remains an open challenge. In this work, we present novel
computational experiments that map and visualize the local mode structure of
Gibbs densities. Efficient mapping requires identifying the global basins
without enumerating the countless modes. Inspired by Grenander's jump-diffusion
method, we propose a new MCMC tool called Attraction-Diffusion (AD) that can
capture the macroscopic structure of highly non-convex densities by measuring
metastability of local modes. AD involves altering the target density with a
magnetization potential penalizing distance from a known mode and running an
MCMC sample of the altered density to measure the stability of the initial
chain state. Using a low-dimensional generator network to facilitate
exploration, we map image spaces with up to 12,288 dimensions (64 $\times$ 64
pixels in RGB). Our work shows: (1) AD can efficiently map highly non-convex
probability densities, (2) metastable regions of pattern probability densities
contain coherent groups of images, and (3) the perceptibility of differences
between training images influences the metastability of image basins.
|
Quantitatively Evaluating GANs With Divergences Proposed for Training | Generative adversarial networks (GANs) have been extremely effective in
approximating complex distributions of high-dimensional, input data samples,
and substantial progress has been made in understanding and improving GAN
performance in terms of both theory and application. However, we currently lack
quantitative methods for model assessment. Because of this, while many GAN
variants are being proposed, we have relatively little understanding of their
relative abilities. In this paper, we evaluate the performance of various types
of GANs using divergence and distance functions typically used only for
training. We observe consistency across the various proposed metrics and,
interestingly, the test-time metrics do not favour networks that use the same
training-time criterion. We also compare the proposed metrics to human
perceptual scores.
|
Specialized Interior Point Algorithm for Stable Nonlinear System
Identification | Estimation of nonlinear dynamic models from data poses many challenges,
including model instability and non-convexity of long-term simulation fidelity.
Recently Lagrangian relaxation has been proposed as a method to approximate
simulation fidelity and guarantee stability via semidefinite programming (SDP),
however the resulting SDPs have large dimension, limiting their utility in
practical problems. In this paper we develop a path-following interior point
algorithm that takes advantage of special structure in the problem and reduces
computational complexity from cubic to linear growth with the length of the
data set. The new algorithm enables empirical comparisons to established
methods including Nonlinear ARX, and we demonstrate superior generalization to
new data. We also explore the "regularizing" effect of stability constraints as
an alternative to regressor subset selection.
|
Practical Contextual Bandits with Regression Oracles | A major challenge in contextual bandits is to design general-purpose
algorithms that are both practically useful and theoretically well-founded. We
present a new technique that has the empirical and computational advantages of
realizability-based approaches combined with the flexibility of agnostic
methods. Our algorithms leverage the availability of a regression oracle for
the value-function class, a more realistic and reasonable oracle than the
classification oracles over policies typically assumed by agnostic methods. Our
approach generalizes both UCB and LinUCB to far more expressive possible model
classes and achieves low regret under certain distributional assumptions. In an
extensive empirical evaluation, compared to both realizability-based and
agnostic baselines, we find that our approach typically gives comparable or
superior results.
|
Slow and Stale Gradients Can Win the Race: Error-Runtime Trade-offs in
Distributed SGD | Distributed Stochastic Gradient Descent (SGD) when run in a synchronous
manner, suffers from delays in waiting for the slowest learners (stragglers).
Asynchronous methods can alleviate stragglers, but cause gradient staleness
that can adversely affect convergence. In this work we present a novel
theoretical characterization of the speed-up offered by asynchronous methods by
analyzing the trade-off between the error in the trained model and the actual
training runtime (wallclock time). The novelty in our work is that our runtime
analysis considers random straggler delays, which helps us design and compare
distributed SGD algorithms that strike a balance between stragglers and
staleness. We also present a new convergence analysis of asynchronous SGD
variants without bounded or exponential delay assumptions, and a novel learning
rate schedule to compensate for gradient staleness.
|
Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with
Adversarial Examples | Crafting adversarial examples has become an important technique to evaluate
the robustness of deep neural networks (DNNs). However, most existing works
focus on attacking the image classification problem since its input space is
continuous and output space is finite.
In this paper, we study the much more challenging problem of crafting
adversarial examples for sequence-to-sequence (seq2seq) models, whose inputs
are discrete text strings and outputs have an almost infinite number of
possibilities. To address the challenges caused by the discrete input space, we
propose a projected gradient method combined with group lasso and gradient
regularization. To handle the almost infinite output space, we design some
novel loss functions to conduct non-overlapping attack and targeted keyword
attack. We apply our algorithm to machine translation and text summarization
tasks, and verify the effectiveness of the proposed algorithm: by changing less
than 3 words, we can make seq2seq model to produce desired outputs with high
success rates. On the other hand, we recognize that, compared with the
well-evaluated CNN-based classifiers, seq2seq models are intrinsically more
robust to adversarial attacks.
|
OIL: Observational Imitation Learning | Recent work has explored the problem of autonomous navigation by imitating a
teacher and learning an end-to-end policy, which directly predicts controls
from raw images. However, these approaches tend to be sensitive to mistakes by
the teacher and do not scale well to other environments or vehicles. To this
end, we propose Observational Imitation Learning (OIL), a novel imitation
learning variant that supports online training and automatic selection of
optimal behavior by observing multiple imperfect teachers. We apply our
proposed methodology to the challenging problems of autonomous driving and UAV
racing. For both tasks, we utilize the Sim4CV simulator that enables the
generation of large amounts of synthetic training data and also allows for
online learning and evaluation. We train a perception network to predict
waypoints from raw image data and use OIL to train another network to predict
controls from these waypoints. Extensive experiments demonstrate that our
trained network outperforms its teachers, conventional imitation learning (IL)
and reinforcement learning (RL) baselines and even humans in simulation. The
project website is available at https://sites.google.com/kaust.edu.sa/oil/ and
a video at https://youtu.be/_rhq8a0qgeg
|
Chest X-Ray Analysis of Tuberculosis by Deep Learning with Segmentation
and Augmentation | The results of chest X-ray (CXR) analysis of 2D images to get the
statistically reliable predictions (availability of tuberculosis) by
computer-aided diagnosis (CADx) on the basis of deep learning are presented.
They demonstrate the efficiency of lung segmentation, lossless and lossy data
augmentation for CADx of tuberculosis by deep convolutional neural network
(CNN) applied to the small and not well-balanced dataset even. CNN demonstrates
ability to train (despite overfitting) on the pre-processed dataset obtained
after lung segmentation in contrast to the original not-segmented dataset.
Lossless data augmentation of the segmented dataset leads to the lowest
validation loss (without overfitting) and nearly the same accuracy (within the
limits of standard deviation) in comparison to the original and other
pre-processed datasets after lossy data augmentation. The additional limited
lossy data augmentation results in the lower validation loss, but with a
decrease of the validation accuracy. In conclusion, besides the more complex
deep CNNs and bigger datasets, the better progress of CADx for the small and
not well-balanced datasets even could be obtained by better segmentation, data
augmentation, dataset stratification, and exclusion of non-evident outliers.
|
Multiresolution Tensor Decomposition for Multiple Spatial Passing
Networks | This article is motivated by soccer positional passing networks collected
across multiple games. We refer to these data as replicated spatial passing
networks---to accurately model such data it is necessary to take into account
the spatial positions of the passer and receiver for each passing event. This
spatial registration and replicates that occur across games represent key
differences with usual social network data. As a key step before investigating
how the passing dynamics influence team performance, we focus on developing
methods for summarizing different team's passing strategies. Our proposed
approach relies on a novel multiresolution data representation framework and
Poisson nonnegative block term decomposition model, which automatically
produces coarse-to-fine low-rank network motifs. The proposed methods are
applied to detailed passing record data collected from the 2014 FIFA World Cup.
|
On the Power of Over-parametrization in Neural Networks with Quadratic
Activation | We provide new theoretical insights on why over-parametrization is effective
in learning neural networks. For a $k$ hidden node shallow network with
quadratic activation and $n$ training data points, we show as long as $ k \ge
\sqrt{2n}$, over-parametrization enables local search algorithms to find a
\emph{globally} optimal solution for general smooth and convex loss functions.
Further, despite that the number of parameters may exceed the sample size,
using theory of Rademacher complexity, we show with weight decay, the solution
also generalizes well if the data is sampled from a regular distribution such
as Gaussian. To prove when $k\ge \sqrt{2n}$, the loss function has benign
landscape properties, we adopt an idea from smoothed analysis, which may have
other applications in studying loss surfaces of neural networks.
|
Deep Bayesian Active Semi-Supervised Learning | In many applications the process of generating label information is expensive
and time consuming. We present a new method that combines active and
semi-supervised deep learning to achieve high generalization performance from a
deep convolutional neural network with as few known labels as possible. In a
setting where a small amount of labeled data as well as a large amount of
unlabeled data is available, our method first learns the labeled data set. This
initialization is followed by an expectation maximization algorithm, where
further training reduces classification entropy on the unlabeled data by
targeting a low entropy fit which is consistent with the labeled data. In
addition the algorithm asks at a specified frequency an oracle for labels of
data with entropy above a certain entropy quantile. Using this active learning
component we obtain an agile labeling process that achieves high accuracy, but
requires only a small amount of known labels. For the MNIST dataset we report
an error rate of 2.06% using only 300 labels and 1.06% for 1000 labels. These
results are obtained without employing any special network architecture or data
augmentation.
|
GAN-based Synthetic Medical Image Augmentation for increased CNN
Performance in Liver Lesion Classification | Deep learning methods, and in particular convolutional neural networks
(CNNs), have led to an enormous breakthrough in a wide range of computer vision
tasks, primarily by using large-scale annotated datasets. However, obtaining
such datasets in the medical domain remains a challenge. In this paper, we
present methods for generating synthetic medical images using recently
presented deep learning Generative Adversarial Networks (GANs). Furthermore, we
show that generated medical images can be used for synthetic data augmentation,
and improve the performance of CNN for medical image classification. Our novel
method is demonstrated on a limited dataset of computed tomography (CT) images
of 182 liver lesions (53 cysts, 64 metastases and 65 hemangiomas). We first
exploit GAN architectures for synthesizing high quality liver lesion ROIs. Then
we present a novel scheme for liver lesion classification using CNN. Finally,
we train the CNN using classic data augmentation and our synthetic data
augmentation and compare performance. In addition, we explore the quality of
our synthesized examples using visualization and expert assessment. The
classification performance using only classic data augmentation yielded 78.6%
sensitivity and 88.4% specificity. By adding the synthetic data augmentation
the results increased to 85.7% sensitivity and 92.4% specificity. We believe
that this approach to synthetic data augmentation can generalize to other
medical classification applications and thus support radiologists' efforts to
improve diagnosis.
|
Fast and Sample Efficient Inductive Matrix Completion via Multi-Phase
Procrustes Flow | We revisit the inductive matrix completion problem that aims to recover a
rank-$r$ matrix with ambient dimension $d$ given $n$ features as the side prior
information. The goal is to make use of the known $n$ features to reduce sample
and computational complexities. We present and analyze a new gradient-based
non-convex optimization algorithm that converges to the true underlying matrix
at a linear rate with sample complexity only linearly depending on $n$ and
logarithmically depending on $d$. To the best of our knowledge, all previous
algorithms either have a quadratic dependency on the number of features in
sample complexity or a sub-linear computational convergence rate. In addition,
we provide experiments on both synthetic and real world data to demonstrate the
effectiveness of our proposed algorithm.
|
Revisiting Spatial-Temporal Similarity: A Deep Learning Framework for
Traffic Prediction | Traffic prediction has drawn increasing attention in AI research field due to
the increasing availability of large-scale traffic data and its importance in
the real world. For example, an accurate taxi demand prediction can assist taxi
companies in pre-allocating taxis. The key challenge of traffic prediction lies
in how to model the complex spatial dependencies and temporal dynamics.
Although both factors have been considered in modeling, existing works make
strong assumptions about spatial dependence and temporal dynamics, i.e.,
spatial dependence is stationary in time, and temporal dynamics is strictly
periodical. However, in practice, the spatial dependence could be dynamic
(i.e., changing from time to time), and the temporal dynamics could have some
perturbation from one period to another period. In this paper, we make two
important observations: (1) the spatial dependencies between locations are
dynamic; and (2) the temporal dependency follows daily and weekly pattern but
it is not strictly periodic for its dynamic temporal shifting. To address these
two issues, we propose a novel Spatial-Temporal Dynamic Network (STDN), in
which a flow gating mechanism is introduced to learn the dynamic similarity
between locations, and a periodically shifted attention mechanism is designed
to handle long-term periodic temporal shifting. To the best of our knowledge,
this is the first work that tackles both issues in a unified framework. Our
experimental results on real-world traffic datasets verify the effectiveness of
the proposed method.
|
Nonnegative Matrix Factorization for Signal and Data Analytics:
Identifiability, Algorithms, and Applications | Nonnegative matrix factorization (NMF) has become a workhorse for signal and
data analytics, triggered by its model parsimony and interpretability. Perhaps
a bit surprisingly, the understanding to its model identifiability---the major
reason behind the interpretability in many applications such as topic mining
and hyperspectral imaging---had been rather limited until recent years.
Beginning from the 2010s, the identifiability research of NMF has progressed
considerably: Many interesting and important results have been discovered by
the signal processing (SP) and machine learning (ML) communities. NMF
identifiability has a great impact on many aspects in practice, such as
ill-posed formulation avoidance and performance-guaranteed algorithm design. On
the other hand, there is no tutorial paper that introduces NMF from an
identifiability viewpoint. In this paper, we aim at filling this gap by
offering a comprehensive and deep tutorial on model identifiability of NMF as
well as the connections to algorithms and applications. This tutorial will help
researchers and graduate students grasp the essence and insights of NMF,
thereby avoiding typical `pitfalls' that are often times due to unidentifiable
NMF formulations. This paper will also help practitioners pick/design suitable
factorization tools for their own problems.
|
An Empirical Evaluation of Generic Convolutional and Recurrent Networks
for Sequence Modeling | For most deep learning practitioners, sequence modeling is synonymous with
recurrent networks. Yet recent results indicate that convolutional
architectures can outperform recurrent networks on tasks such as audio
synthesis and machine translation. Given a new sequence modeling task or
dataset, which architecture should one use? We conduct a systematic evaluation
of generic convolutional and recurrent architectures for sequence modeling. The
models are evaluated across a broad range of standard tasks that are commonly
used to benchmark recurrent networks. Our results indicate that a simple
convolutional architecture outperforms canonical recurrent networks such as
LSTMs across a diverse range of tasks and datasets, while demonstrating longer
effective memory. We conclude that the common association between sequence
modeling and recurrent networks should be reconsidered, and convolutional
networks should be regarded as a natural starting point for sequence modeling
tasks. To assist related work, we have made code available at
http://github.com/locuslab/TCN .
|
Accelerating Natural Gradient with Higher-Order Invariance | An appealing property of the natural gradient is that it is invariant to
arbitrary differentiable reparameterizations of the model. However, this
invariance property requires infinitesimal steps and is lost in practical
implementations with small but finite step sizes. In this paper, we study
invariance properties from a combined perspective of Riemannian geometry and
numerical differential equation solving. We define the order of invariance of a
numerical method to be its convergence order to an invariant solution. We
propose to use higher-order integrators and geodesic corrections to obtain more
invariant optimization trajectories. We prove the numerical convergence
properties of geodesic corrected updates and show that they can be as
computationally efficient as plain natural gradient. Experimentally, we
demonstrate that invariance leads to faster optimization and our techniques
improve on traditional natural gradient in deep neural network training and
natural policy gradient for reinforcement learning.
|
An Optimal Control Approach to Deep Learning and Applications to
Discrete-Weight Neural Networks | Deep learning is formulated as a discrete-time optimal control problem. This
allows one to characterize necessary conditions for optimality and develop
training algorithms that do not rely on gradients with respect to the trainable
parameters. In particular, we introduce the discrete-time method of successive
approximations (MSA), which is based on the Pontryagin's maximum principle, for
training neural networks. A rigorous error estimate for the discrete MSA is
obtained, which sheds light on its dynamics and the means to stabilize the
algorithm. The developed methods are applied to train, in a rather principled
way, neural networks with weights that are constrained to take values in a
discrete set. We obtain competitive performance and interestingly, very sparse
weights in the case of ternary networks, which may be useful in model
deployment in low-memory devices.
|
Distributed Nonparametric Regression under Communication Constraints | This paper studies the problem of nonparametric estimation of a smooth
function with data distributed across multiple machines. We assume an
independent sample from a white noise model is collected at each machine, and
an estimator of the underlying true function needs to be constructed at a
central machine. We place limits on the number of bits that each machine can
use to transmit information to the central machine. Our results give both
asymptotic lower bounds and matching upper bounds on the statistical risk under
various settings. We identify three regimes, depending on the relationship
among the number of machines, the size of the data available at each machine,
and the communication budget. When the communication budget is small, the
statistical risk depends solely on this communication bottleneck, regardless of
the sample size. In the regime where the communication budget is large, the
classic minimax risk in the non-distributed estimation setting is recovered. In
an intermediate regime, the statistical risk depends on both the sample size
and the communication budget.
|
On Cognitive Preferences and the Plausibility of Rule-based Models | It is conventional wisdom in machine learning and data mining that logical
models such as rule sets are more interpretable than other models, and that
among such rule-based models, simpler models are more interpretable than more
complex ones. In this position paper, we question this latter assumption by
focusing on one particular aspect of interpretability, namely the plausibility
of models. Roughly speaking, we equate the plausibility of a model with the
likeliness that a user accepts it as an explanation for a prediction. In
particular, we argue that, all other things being equal, longer explanations
may be more convincing than shorter ones, and that the predominant bias for
shorter models, which is typically necessary for learning powerful
discriminative models, may not be suitable when it comes to user acceptance of
the learned models. To that end, we first recapitulate evidence for and against
this postulate, and then report the results of an evaluation in a
crowd-sourcing study based on about 3.000 judgments. The results do not reveal
a strong preference for simple rules, whereas we can observe a weak preference
for longer rules in some domains. We then relate these results to well-known
cognitive biases such as the conjunction fallacy, the representative heuristic,
or the recogition heuristic, and investigate their relation to rule length and
plausibility.
|
Greedy stochastic algorithms for entropy-regularized optimal transport
problems | Optimal transport (OT) distances are finding evermore applications in machine
learning and computer vision, but their wide spread use in larger-scale
problems is impeded by their high computational cost. In this work we develop a
family of fast and practical stochastic algorithms for solving the optimal
transport problem with an entropic penalization. This work extends the recently
developed Greenkhorn algorithm, in the sense that, the Greenkhorn algorithm is
a limiting case of this family. We also provide a simple and general
convergence theorem for all algorithms in the class, with rates that match the
best known rates of Greenkorn and the Sinkhorn algorithm, and conclude with
numerical experiments that show under what regime of penalization the new
stochastic methods are faster than the aforementioned methods.
|
Deep Network Regularization via Bayesian Inference of Synaptic
Connectivity | Deep neural networks (DNNs) often require good regularizers to generalize
well. Currently, state-of-the-art DNN regularization techniques consist in
randomly dropping units and/or connections on each iteration of the training
algorithm. Dropout and DropConnect are characteristic examples of such
regularizers, that are widely popular among practitioners. However, a drawback
of such approaches consists in the fact that their postulated probability of
random unit/connection omission is a constant that must be heuristically
selected based on the obtained performance in some validation set. To alleviate
this burden, in this paper we regard the DNN regularization problem from a
Bayesian inference perspective: We impose a sparsity-inducing prior over the
network synaptic weights, where the sparsity is induced by a set of
Bernoulli-distributed binary variables with Beta (hyper-)priors over their
prior parameters. This way, we eventually allow for marginalizing over the DNN
synaptic connectivity for output generation, thus giving rise to an effective,
heuristics-free, network regularization scheme. We perform Bayesian inference
for the resulting hierarchical model by means of an efficient Black-Box
Variational inference scheme. We exhibit the advantages of our method over
existing approaches by conducting an extensive experimental evaluation using
benchmark datasets.
|
SAFE: Spectral Evolution Analysis Feature Extraction for Non-Stationary
Time Series Prediction | This paper presents a practical approach for detecting non-stationarity in
time series prediction. This method is called SAFE and works by monitoring the
evolution of the spectral contents of time series through a distance function.
This method is designed to work in combination with state-of-the-art machine
learning methods in real time by informing the online predictors to perform
necessary adaptation when a non-stationarity presents. We also propose an
algorithm to proportionally include some past data in the adaption process to
overcome the Catastrophic Forgetting problem. To validate our hypothesis and
test the effectiveness of our approach, we present comprehensive experiments in
different elements of the approach involving artificial and real-world
datasets. The experiments show that the proposed method is able to
significantly save computational resources in term of processor or GPU cycles
while maintaining high prediction performances.
|
New Results on Multi-Step Traffic Flow Prediction | In its simplest form, the traffic flow prediction problem is restricted to
predicting a single time-step into the future. Multi-step traffic flow
prediction extends this set-up to the case where predicting multiple time-steps
into the future based on some finite history is of interest. This problem is
significantly more difficult than its single-step variant and is known to
suffer from degradation in predictions as the time step increases. In this
paper, two approaches to improve multi-step traffic flow prediction performance
in recursive and multi-output settings are introduced. In particular, a model
that allows recursive prediction approaches to take into account the temporal
context in term of time-step index when making predictions is introduced. In
addition, a conditional generative adversarial network-based data augmentation
method is proposed to improve prediction performance in the multi-output
setting. The experiments on a real-world traffic flow dataset show that the two
methods improve on multi-step traffic flow prediction in recursive and
multi-output settings, respectively.
|
A Distributed Quasi-Newton Algorithm for Empirical Risk Minimization
with Nonsmooth Regularization | We propose a communication- and computation-efficient distributed
optimization algorithm using second-order information for solving ERM problems
with a nonsmooth regularization term. Current second-order and quasi-Newton
methods for this problem either do not work well in the distributed setting or
work only for specific regularizers. Our algorithm uses successive quadratic
approximations, and we describe how to maintain an approximation of the Hessian
and solve subproblems efficiently in a distributed manner. The proposed method
enjoys global linear convergence for a broad range of non-strongly convex
problems that includes the most commonly used ERMs, thus requiring lower
communication complexity. It also converges on non-convex problems, so has the
potential to be used on applications such as deep learning. Initial
computational results on convex problems demonstrate that our method
significantly improves on communication cost and running time over the current
state-of-the-art methods.
|
Detecting Correlations with Little Memory and Communication | We study the problem of identifying correlations in multivariate data, under
information constraints: Either on the amount of memory that can be used by the
algorithm, or the amount of communication when the data is distributed across
several machines. We prove a tight trade-off between the memory/communication
complexity and the sample complexity, implying (for example) that to detect
pairwise correlations with optimal sample complexity, the number of required
memory/communication bits is at least quadratic in the dimension. Our results
substantially improve those of Shamir [2014], which studied a similar question
in a much more restricted setting. To the best of our knowledge, these are the
first provable sample/memory/communication trade-offs for a practical
estimation problem, using standard distributions, and in the natural regime
where the memory/communication budget is larger than the size of a single data
point. To derive our theorems, we prove a new information-theoretic result,
which may be relevant for studying other information-constrained learning
problems.
|
DAGs with NO TEARS: Continuous Optimization for Structure Learning | Estimating the structure of directed acyclic graphs (DAGs, also known as
Bayesian networks) is a challenging problem since the search space of DAGs is
combinatorial and scales superexponentially with the number of nodes. Existing
approaches rely on various local heuristics for enforcing the acyclicity
constraint. In this paper, we introduce a fundamentally different strategy: We
formulate the structure learning problem as a purely \emph{continuous}
optimization problem over real matrices that avoids this combinatorial
constraint entirely. This is achieved by a novel characterization of acyclicity
that is not only smooth but also exact. The resulting problem can be
efficiently solved by standard numerical algorithms, which also makes
implementation effortless. The proposed method outperforms existing ones,
without imposing any structural assumptions on the graph such as bounded
treewidth or in-degree. Code implementing the proposed algorithm is open-source
and publicly available at https://github.com/xunzheng/notears.
|
Hierarchical Modeling and Shrinkage for User Session Length Prediction
in Media Streaming | An important metric of users' satisfaction and engagement within on-line
streaming services is the user session length, i.e. the amount of time they
spend on a service continuously without interruption. Being able to predict
this value directly benefits the recommendation and ad pacing contexts in music
and video streaming services. Recent research has shown that predicting the
exact amount of time spent is highly nontrivial due to many external factors
for which a user can end a session, and the lack of predictive covariates. Most
of the other related literature on duration based user engagement has focused
on dwell time for websites, for search and display ads, mainly for post-click
satisfaction prediction or ad ranking.
In this work we present a novel framework inspired by hierarchical Bayesian
modeling to predict, at the moment of login, the amount of time a user will
spend in the streaming service. The time spent by a user on a platform depends
upon user-specific latent variables which are learned via hierarchical
shrinkage. Our framework enjoys theoretical guarantees and naturally
incorporates flexible parametric/nonparametric models on the covariates,
including models robust to outliers. Our proposal is found to outperform
state-of- the-art estimators in terms of efficiency and predictive performance
on real world public and private datasets.
|
Stochastic Activation Pruning for Robust Adversarial Defense | Neural networks are known to be vulnerable to adversarial examples. Carefully
chosen perturbations to real images, while imperceptible to humans, induce
misclassification and threaten the reliability of deep learning systems in the
wild. To guard against adversarial examples, we take inspiration from game
theory and cast the problem as a minimax zero-sum game between the adversary
and the model. In general, for such games, the optimal strategy for both
players requires a stochastic policy, also known as a mixed strategy. In this
light, we propose Stochastic Activation Pruning (SAP), a mixed strategy for
adversarial defense. SAP prunes a random subset of activations (preferentially
pruning those with smaller magnitude) and scales up the survivors to
compensate. We can apply SAP to pretrained networks, including adversarially
trained models, without fine-tuning, providing robustness against adversarial
examples. Experiments demonstrate that SAP confers robustness against attacks,
increasing accuracy and preserving calibration.
|
Deep Continuous Clustering | Clustering high-dimensional datasets is hard because interpoint distances
become less informative in high-dimensional spaces. We present a clustering
algorithm that performs nonlinear dimensionality reduction and clustering
jointly. The data is embedded into a lower-dimensional space by a deep
autoencoder. The autoencoder is optimized as part of the clustering process.
The resulting network produces clustered data. The presented approach does not
rely on prior knowledge of the number of ground-truth clusters. Joint nonlinear
dimensionality reduction and clustering are formulated as optimization of a
global continuous objective. We thus avoid discrete reconfigurations of the
objective that characterize prior clustering algorithms. Experiments on
datasets from multiple domains demonstrate that the presented algorithm
outperforms state-of-the-art clustering schemes, including recent methods that
use deep networks.
|
Query and Output: Generating Words by Querying Distributed Word
Representations for Paraphrase Generation | Most recent approaches use the sequence-to-sequence model for paraphrase
generation. The existing sequence-to-sequence model tends to memorize the words
and the patterns in the training dataset instead of learning the meaning of the
words. Therefore, the generated sentences are often grammatically correct but
semantically improper. In this work, we introduce a novel model based on the
encoder-decoder framework, called Word Embedding Attention Network (WEAN). Our
proposed model generates the words by querying distributed word representations
(i.e. neural word embeddings), hoping to capturing the meaning of the according
words. Following previous work, we evaluate our model on two
paraphrase-oriented tasks, namely text simplification and short text
abstractive summarization. Experimental results show that our model outperforms
the sequence-to-sequence baseline by the BLEU score of 6.3 and 5.5 on two
English text simplification datasets, and the ROUGE-2 F1 score of 5.7 on a
Chinese summarization dataset. Moreover, our model achieves state-of-the-art
performances on these three benchmark datasets.
|
Totally Looks Like - How Humans Compare, Compared to Machines | Perceptual judgment of image similarity by humans relies on rich internal
representations ranging from low-level features to high-level concepts, scene
properties and even cultural associations. However, existing methods and
datasets attempting to explain perceived similarity use stimuli which arguably
do not cover the full breadth of factors that affect human similarity
judgments, even those geared toward this goal. We introduce a new dataset
dubbed Totally-Looks-Like (TLL) after a popular entertainment website, which
contains images paired by humans as being visually similar. The dataset
contains 6016 image-pairs from the wild, shedding light upon a rich and diverse
set of criteria employed by human beings. We conduct experiments to try to
reproduce the pairings via features extracted from state-of-the-art deep
convolutional neural networks, as well as additional human experiments to
verify the consistency of the collected data. Though we create conditions to
artificially make the matching task increasingly easier, we show that
machine-extracted representations perform very poorly in terms of reproducing
the matching selected by humans. We discuss and analyze these results,
suggesting future directions for improvement of learned image representations.
|
Recurrent Predictive State Policy Networks | We introduce Recurrent Predictive State Policy (RPSP) networks, a recurrent
architecture that brings insights from predictive state representations to
reinforcement learning in partially observable environments. Predictive state
policy networks consist of a recursive filter, which keeps track of a belief
about the state of the environment, and a reactive policy that directly maps
beliefs to actions, to maximize the cumulative reward. The recursive filter
leverages predictive state representations (PSRs) (Rosencrantz and Gordon,
2004; Sun et al., 2016) by modeling predictive state-- a prediction of the
distribution of future observations conditioned on history and future actions.
This representation gives rise to a rich class of statistically consistent
algorithms (Hefny et al., 2018) to initialize the recursive filter. Predictive
state serves as an equivalent representation of a belief state. Therefore, the
policy component of the RPSP-network can be purely reactive, simplifying
training while still allowing optimal behaviour. Moreover, we use the PSR
interpretation during training as well, by incorporating prediction error in
the loss function. The entire network (recursive filter and reactive policy) is
still differentiable and can be trained using gradient based methods. We
optimize our policy using a combination of policy gradient based on rewards
(Williams, 1992) and gradient descent based on prediction error. We show the
efficacy of RPSP-networks under partial observability on a set of robotic
control tasks from OpenAI Gym. We empirically show that RPSP-networks perform
well compared with memory-preserving networks such as GRUs, as well as finite
memory models, being the overall best performing method.
|
Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates | In large-scale distributed learning, security issues have become increasingly
important. Particularly in a decentralized environment, some computing units
may behave abnormally, or even exhibit Byzantine failures -- arbitrary and
potentially adversarial behavior. In this paper, we develop distributed
learning algorithms that are provably robust against such failures, with a
focus on achieving optimal statistical performance. A main result of this work
is a sharp analysis of two robust distributed gradient descent algorithms based
on median and trimmed mean operations, respectively. We prove statistical error
rates for three kinds of population loss functions: strongly convex,
non-strongly convex, and smooth non-convex. In particular, these algorithms are
shown to achieve order-optimal statistical error rates for strongly convex
losses. To achieve better communication efficiency, we further propose a
median-based distributed algorithm that is provably robust, and uses only one
communication round. For strongly convex quadratic loss, we show that this
algorithm achieves the same optimal error rate as the robust distributed
gradient descent algorithms.
|
Memorization Precedes Generation: Learning Unsupervised GANs with Memory
Networks | We propose an approach to address two issues that commonly occur during
training of unsupervised GANs. First, since GANs use only a continuous latent
distribution to embed multiple classes or clusters of data, they often do not
correctly handle the structural discontinuity between disparate classes in a
latent space. Second, discriminators of GANs easily forget about past generated
samples by generators, incurring instability during adversarial training. We
argue that these two infamous problems of unsupervised GAN training can be
largely alleviated by a learnable memory network to which both generators and
discriminators can access. Generators can effectively learn representation of
training samples to understand underlying cluster distributions of data, which
ease the structure discontinuity problem. At the same time, discriminators can
better memorize clusters of previously generated samples, which mitigate the
forgetting problem. We propose a novel end-to-end GAN model named memoryGAN,
which involves a memory network that is unsupervisedly trainable and integrable
to many existing GAN models. With evaluations on multiple datasets such as
Fashion-MNIST, CelebA, CIFAR10, and Chairs, we show that our model is
probabilistically interpretable, and generates realistic image samples of high
visual fidelity. The memoryGAN also achieves the state-of-the-art inception
scores over unsupervised GAN models on the CIFAR10 dataset, without any
optimization tricks and weaker divergences.
|
Blind Channel Equalization using Variational Autoencoders | A new maximum likelihood estimation approach for blind channel equalization,
using variational autoencoders (VAEs), is introduced. Significant and
consistent improvements in the error rate of the reconstructed symbols,
compared to constant modulus equalizers, are demonstrated. In fact, for the
channels that were examined, the performance of the new VAE blind channel
equalizer was close to the performance of a nonblind adaptive linear minimum
mean square error equalizer. The new equalization method enables a
significantly lower latency channel acquisition compared to the constant
modulus algorithm (CMA). The VAE uses a convolutional neural network with two
layers and a very small number of free parameters. Although the computational
complexity of the new equalizer is higher compared to CMA, it is still
reasonable, and the number of free parameters to estimate is small.
|
Improving the Improved Training of Wasserstein GANs: A Consistency Term
and Its Dual Effect | Despite being impactful on a variety of problems and applications, the
generative adversarial nets (GANs) are remarkably difficult to train. This
issue is formally analyzed by \cite{arjovsky2017towards}, who also propose an
alternative direction to avoid the caveats in the minmax two-player training of
GANs. The corresponding algorithm, called Wasserstein GAN (WGAN), hinges on the
1-Lipschitz continuity of the discriminator. In this paper, we propose a novel
approach to enforcing the Lipschitz continuity in the training procedure of
WGANs. Our approach seamlessly connects WGAN with one of the recent
semi-supervised learning methods. As a result, it gives rise to not only better
photo-realistic samples than the previous methods but also state-of-the-art
semi-supervised learning results. In particular, our approach gives rise to the
inception score of more than 5.0 with only 1,000 CIFAR-10 images and is the
first that exceeds the accuracy of 90% on the CIFAR-10 dataset using only 4,000
labeled images, to the best of our knowledge.
|
Online learning over a finite action set with limited switching | This paper studies the value of switching actions in the Prediction From
Experts (PFE) problem and Adversarial Multi-Armed Bandits (MAB) problem. First,
we revisit the well-studied and practically motivated setting of PFE with
switching costs. Many algorithms are known to achieve the minimax optimal order
of $O(\sqrt{T \log n})$ in expectation for both regret and number of switches,
where $T$ is the number of iterations and $n$ the number of actions. However,
no high probability (h.p.) guarantees are known. Our main technical
contribution is the first algorithms which with h.p. achieve this optimal order
for both regret and switches. This settles an open problem of [Devroye et al.,
2015], and directly implies the first h.p. guarantees for several problems of
interest.
Next, to investigate the value of switching actions at a more granular level,
we introduce the setting of switching budgets, in which algorithms are limited
to $S \leq T$ switches between actions. This entails a limited number of free
switches, in contrast to the unlimited number of expensive switches in the
switching cost setting. Using the above result and several reductions, we unify
previous work and completely characterize the complexity of this switching
budget setting up to small polylogarithmic factors: for both PFE and MAB, for
all switching budgets $S \leq T$, and for both expectation and h.p. guarantees.
For PFE, we show the optimal rate is $\tilde{\Theta}(\sqrt{T\log n})$ for $S =
\Omega(\sqrt{T\log n})$, and $\min(\tilde{\Theta}(\tfrac{T\log n}{S}), T)$ for
$S = O(\sqrt{T \log n})$. Interestingly, the bandit setting does not exhibit
such a phase transition; instead we show the minimax rate decays steadily as
$\min(\tilde{\Theta}(\tfrac{T\sqrt{n}}{\sqrt{S}}), T)$ for all ranges of $S
\leq T$. These results recover and generalize the known minimax rates for the
(arbitrary) switching cost setting.
|
Local Distance Metric Learning for Nearest Neighbor Algorithm | Distance metric learning is a successful way to enhance the performance of
the nearest neighbor classifier. In most cases, however, the distribution of
data does not obey a regular form and may change in different parts of the
feature space. Regarding that, this paper proposes a novel local distance
metric learning method, namely Local Mahalanobis Distance Learning (LMDL), in
order to enhance the performance of the nearest neighbor classifier. LMDL
considers the neighborhood influence and learns multiple distance metrics for a
reduced set of input samples. The reduced set is called as prototypes which try
to preserve local discriminative information as much as possible. The proposed
LMDL can be kernelized very easily, which is significantly desirable in the
case of highly nonlinear data. The quality as well as the efficiency of the
proposed method assesses through a set of different experiments on various
datasets and the obtained results show that LDML as well as the kernelized
version is superior to the other related state-of-the-art methods.
|
Adversarial Extreme Multi-label Classification | The goal in extreme multi-label classification is to learn a classifier which
can assign a small subset of relevant labels to an instance from an extremely
large set of target labels. Datasets in extreme classification exhibit a long
tail of labels which have small number of positive training instances. In this
work, we pose the learning task in extreme classification with large number of
tail-labels as learning in the presence of adversarial perturbations. This view
motivates a robust optimization framework and equivalence to a corresponding
regularized objective.
Under the proposed robustness framework, we demonstrate efficacy of Hamming
loss for tail-label detection in extreme classification. The equivalent
regularized objective, in combination with proximal gradient based
optimization, performs better than state-of-the-art methods on propensity
scored versions of precision@k and nDCG@k(upto 20% relative improvement over
PFastreXML - a leading tree-based approach and 60% relative improvement over
SLEEC - a leading label-embedding approach). Furthermore, we also highlight the
sub-optimality of a sparse solver in a widely used package for large-scale
linear classification, which is interesting in its own right. We also
investigate the spectral properties of label graphs for providing novel
insights towards understanding the conditions governing the performance of
Hamming loss based one-vs-rest scheme vis-\`a-vis label embedding methods.
|
A Comparative Study of Pairwise Learning Methods based on Kernel Ridge
Regression | Many machine learning problems can be formulated as predicting labels for a
pair of objects. Problems of that kind are often referred to as pairwise
learning, dyadic prediction or network inference problems. During the last
decade kernel methods have played a dominant role in pairwise learning. They
still obtain a state-of-the-art predictive performance, but a theoretical
analysis of their behavior has been underexplored in the machine learning
literature.
In this work we review and unify existing kernel-based algorithms that are
commonly used in different pairwise learning settings, ranging from matrix
filtering to zero-shot learning. To this end, we focus on closed-form efficient
instantiations of Kronecker kernel ridge regression. We show that independent
task kernel ridge regression, two-step kernel ridge regression and a linear
matrix filter arise naturally as a special case of Kronecker kernel ridge
regression, implying that all these methods implicitly minimize a squared loss.
In addition, we analyze universality, consistency and spectral filtering
properties. Our theoretical results provide valuable insights in assessing the
advantages and limitations of existing pairwise learning methods.
|
N-body Networks: a Covariant Hierarchical Neural Network Architecture
for Learning Atomic Potentials | We describe N-body networks, a neural network architecture for learning the
behavior and properties of complex many body physical systems. Our specific
application is to learn atomic potential energy surfaces for use in molecular
dynamics simulations. Our architecture is novel in that (a) it is based on a
hierarchical decomposition of the many body system into subsytems, (b) the
activations of the network correspond to the internal state of each subsystem,
(c) the "neurons" in the network are constructed explicitly so as to guarantee
that each of the activations is covariant to rotations, (d) the neurons operate
entirely in Fourier space, and the nonlinearities are realized by tensor
products followed by Clebsch-Gordan decompositions. As part of the description
of our network, we give a characterization of what way the weights of the
network may interact with the activations so as to ensure that the covariance
property is maintained.
|
Variance-Aware Regret Bounds for Undiscounted Reinforcement Learning in
MDPs | The problem of reinforcement learning in an unknown and discrete Markov
Decision Process (MDP) under the average-reward criterion is considered, when
the learner interacts with the system in a single stream of observations,
starting from an initial state without any reset. We revisit the minimax lower
bound for that problem by making appear the local variance of the bias function
in place of the diameter of the MDP. Furthermore, we provide a novel analysis
of the KL-UCRL algorithm establishing a high-probability regret bound scaling
as $\widetilde {\mathcal O}\Bigl({\textstyle \sqrt{S\sum_{s,a}{\bf
V}^\star_{s,a}T}}\Big)$ for this algorithm for ergodic MDPs, where $S$ denotes
the number of states and where ${\bf V}^\star_{s,a}$ is the variance of the
bias function with respect to the next-state distribution following action $a$
in state $s$. The resulting bound improves upon the best previously known
regret bound $\widetilde {\mathcal O}(DS\sqrt{AT})$ for that algorithm, where
$A$ and $D$ respectively denote the maximum number of actions (per state) and
the diameter of MDP. We finally compare the leading terms of the two bounds in
some benchmark MDPs indicating that the derived bound can provide an order of
magnitude improvement in some cases. Our analysis leverages novel variations of
the transportation lemma combined with Kullback-Leibler concentration
inequalities, that we believe to be of independent interest.
|
Beyond Greedy Ranking: Slate Optimization via List-CVAE | The conventional solution to the recommendation problem greedily ranks
individual document candidates by prediction scores. However, this method fails
to optimize the slate as a whole, and hence, often struggles to capture biases
caused by the page layout and document interdepedencies. The slate
recommendation problem aims to directly find the optimally ordered subset of
documents (i.e. slates) that best serve users' interests. Solving this problem
is hard due to the combinatorial explosion in all combinations of document
candidates and their display positions on the page. Therefore we propose a
paradigm shift from the traditional viewpoint of solving a ranking problem to a
direct slate generation framework. In this paper, we introduce List Conditional
Variational Auto-Encoders (List-CVAE), which learns the joint distribution of
documents on the slate conditioned on user responses, and directly generates
full slates. Experiments on simulated and real-world data show that List-CVAE
outperforms popular comparable ranking methods consistently on various scales
of documents corpora.
|
On Extended Long Short-term Memory and Dependent Bidirectional Recurrent
Neural Network | In this work, we first analyze the memory behavior in three recurrent neural
networks (RNN) cells; namely, the simple RNN (SRN), the long short-term memory
(LSTM) and the gated recurrent unit (GRU), where the memory is defined as a
function that maps previous elements in a sequence to the current output. Our
study shows that all three of them suffer rapid memory decay. Then, to
alleviate this effect, we introduce trainable scaling factors that act like an
attention mechanism to adjust memory decay adaptively. The new design is called
the extended LSTM (ELSTM). Finally, to design a system that is robust to
previous erroneous predictions, we propose a dependent bidirectional recurrent
neural network (DBRNN). Extensive experiments are conducted on different
language tasks to demonstrate the superiority of the proposed ELSTM and DBRNN
solutions. The ELTSM has achieved up to 30% increase in the labeled attachment
score (LAS) as compared to LSTM and GRU in the dependency parsing (DP) task.
Our models also outperform other state-of-the-art models such as bi-attention
and convolutional sequence to sequence (convseq2seq) by close to 10% in the
LAS. The code is released as an open source
(https://github.com/yuanhangsu/ELSTM-DBRNN)
|
How to Start Training: The Effect of Initialization and Architecture | We identify and study two common failure modes for early training in deep
ReLU nets. For each we give a rigorous proof of when it occurs and how to avoid
it, for fully connected and residual architectures. The first failure mode,
exploding/vanishing mean activation length, can be avoided by initializing
weights from a symmetric distribution with variance 2/fan-in and, for ResNets,
by correctly weighting the residual modules. We prove that the second failure
mode, exponentially large variance of activation length, never occurs in
residual nets once the first failure mode is avoided. In contrast, for fully
connected nets, we prove that this failure mode can happen and is avoided by
keeping constant the sum of the reciprocals of layer widths. We demonstrate
empirically the effectiveness of our theoretical results in predicting when
networks are able to start training. In particular, we note that many popular
initializations fail our criteria, whereas correct initialization and
architecture allows much deeper networks to be trained.
|
An Analysis of the t-SNE Algorithm for Data Visualization | A first line of attack in exploratory data analysis is data visualization,
i.e., generating a 2-dimensional representation of data that makes clusters of
similar points visually identifiable. Standard Johnson-Lindenstrauss
dimensionality reduction does not produce data visualizations. The t-SNE
heuristic of van der Maaten and Hinton, which is based on non-convex
optimization, has become the de facto standard for visualization in a wide
range of applications.
This work gives a formal framework for the problem of data visualization -
finding a 2-dimensional embedding of clusterable data that correctly separates
individual clusters to make them visually identifiable. We then give a rigorous
analysis of the performance of t-SNE under a natural, deterministic condition
on the "ground-truth" clusters (similar to conditions assumed in earlier
analyses of clustering) in the underlying data. These are the first provable
guarantees on t-SNE for constructing good data visualizations.
We show that our deterministic condition is satisfied by considerably general
probabilistic generative models for clusterable data such as mixtures of
well-separated log-concave distributions. Finally, we give theoretical evidence
that t-SNE provably succeeds in partially recovering cluster structure even
when the above deterministic condition is not met.
|
Kinematic Morphing Networks for Manipulation Skill Transfer | The transfer of a robot skill between different geometric environments is
non-trivial since a wide variety of environments exists, sensor observations as
well as robot motions are high-dimensional, and the environment might only be
partially observed. We consider the problem of extracting a low-dimensional
description of the manipulated environment in form of a kinematic model. This
allows us to transfer a skill by defining a policy on a prototype model and
morphing the observed environment to this prototype. A deep neural network is
used to map depth image observations of the environment to morphing parameter,
which include transformation and configuration parameters of the prototype
model. Using the concatenation property of affine transformations and the
ability to convert point clouds to depth images allows to apply the network in
an iterative manner. The network is trained on data generated in a simulator
and on augmented data that is created by using network predictions. The
algorithm is evaluated on different tasks, where it is shown that iterative
predictions lead to a higher accuracy than one-step predictions.
|
Differentiable Submodular Maximization | We consider learning of submodular functions from data. These functions are
important in machine learning and have a wide range of applications, e.g. data
summarization, feature selection and active learning. Despite their
combinatorial nature, submodular functions can be maximized approximately with
strong theoretical guarantees in polynomial time. Typically, learning the
submodular function and optimization of that function are treated separately,
i.e. the function is first learned using a proxy objective and subsequently
maximized. In contrast, we show how to perform learning and optimization
jointly. By interpreting the output of greedy maximization algorithms as
distributions over sequences of items and smoothening these distributions, we
obtain a differentiable objective. In this way, we can differentiate through
the maximization algorithms and optimize the model to work well with the
optimization algorithm. We theoretically characterize the error made by our
approach, yielding insights into the tradeoff of smoothness and accuracy. We
demonstrate the effectiveness of our approach for jointly learning and
optimizing on synthetic maximum cut data, and on real world applications such
as product recommendation and image collection summarization.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.