title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Online Machine Learning in Big Data Streams | The area of online machine learning in big data streams covers algorithms
that are (1) distributed and (2) work from data streams with only a limited
possibility to store past data. The first requirement mostly concerns software
architectures and efficient algorithms. The second one also imposes nontrivial
theoretical restrictions on the modeling methods: In the data stream model,
older data is no longer available to revise earlier suboptimal modeling
decisions as the fresh data arrives.
In this article, we provide an overview of distributed software architectures
and libraries as well as machine learning models for online learning. We
highlight the most important ideas for classification, regression,
recommendation, and unsupervised modeling from streaming data, and we show how
they are implemented in various distributed data stream processing systems.
This article is a reference material and not a survey. We do not attempt to
be comprehensive in describing all existing methods and solutions; rather, we
give pointers to the most important resources in the field. All related
sub-fields, online algorithms, online learning, and distributed data processing
are hugely dominant in current research and development with conceptually new
research results and software components emerging at the time of writing. In
this article, we refer to several survey results, both for distributed data
processing and for online machine learning. Compared to past surveys, our
article is different because we discuss recommender systems in extended detail.
|
Constrained Convolutional-Recurrent Networks to Improve Speech Quality
with Low Impact on Recognition Accuracy | For a speech-enhancement algorithm, it is highly desirable to simultaneously
improve perceptual quality and recognition rate. Thanks to computational costs
and model complexities, it is challenging to train a model that effectively
optimizes both metrics at the same time. In this paper, we propose a method for
speech enhancement that combines local and global contextual structures
information through convolutional-recurrent neural networks that improves
perceptual quality. At the same time, we introduce a new constraint on the
objective function using a language model/decoder that limits the impact on
recognition rate. Based on experiments conducted with real user data, we
demonstrate that our new context-augmented machine-learning approach for speech
enhancement improves PESQ and WER by an additional 24.5% and 51.3%,
respectively, when compared to the best-performing methods in the literature.
|
Combining Linear Non-Gaussian Acyclic Model with Logistic Regression
Model for Estimating Causal Structure from Mixed Continuous and Discrete Data | Estimating causal models from observational data is a crucial task in data
analysis. For continuous-valued data, Shimizu et al. have proposed a linear
acyclic non-Gaussian model to understand the data generating process, and have
shown that their model is identifiable when the number of data is sufficiently
large. However, situations in which continuous and discrete variables coexist
in the same problem are common in practice. Most existing causal discovery
methods either ignore the discrete data and apply a continuous-valued algorithm
or discretize all the continuous data and then apply a discrete Bayesian
network approach. These methods possibly loss important information when we
ignore discrete data or introduce the approximation error due to
discretization. In this paper, we define a novel hybrid causal model which
consists of both continuous and discrete variables. The model assumes: (1) the
value of a continuous variable is a linear function of its parent variables
plus a non-Gaussian noise, and (2) each discrete variable is a logistic
variable whose distribution parameters depend on the values of its parent
variables. In addition, we derive the BIC scoring function for model selection.
The new discovery algorithm can learn causal structures from mixed continuous
and discrete data without discretization. We empirically demonstrate the power
of our method through thorough simulations.
|
Pattern Localization in Time Series through Signal-To-Model Alignment in
Latent Space | In this paper, we study the problem of locating a predefined sequence of
patterns in a time series. In particular, the studied scenario assumes a
theoretical model is available that contains the expected locations of the
patterns. This problem is found in several contexts, and it is commonly solved
by first synthesizing a time series from the model, and then aligning it to the
true time series through dynamic time warping. We propose a technique that
increases the similarity of both time series before aligning them, by mapping
them into a latent correlation space. The mapping is learned from the data
through a machine-learning setup. Experiments on data from non-destructive
testing demonstrate that the proposed approach shows significant improvements
over the state of the art.
|
Spectral Normalization for Generative Adversarial Networks | One of the challenges in the study of generative adversarial networks is the
instability of its training. In this paper, we propose a novel weight
normalization technique called spectral normalization to stabilize the training
of the discriminator. Our new normalization technique is computationally light
and easy to incorporate into existing implementations. We tested the efficacy
of spectral normalization on CIFAR10, STL-10, and ILSVRC2012 dataset, and we
experimentally confirmed that spectrally normalized GANs (SN-GANs) is capable
of generating images of better or equal quality relative to the previous
training stabilization techniques.
|
WHInter: A Working set algorithm for High-dimensional sparse second
order Interaction models | Learning sparse linear models with two-way interactions is desirable in many
application domains such as genomics. l1-regularised linear models are popular
to estimate sparse models, yet standard implementations fail to address
specifically the quadratic explosion of candidate two-way interactions in high
dimensions, and typically do not scale to genetic data with hundreds of
thousands of features. Here we present WHInter, a working set algorithm to
solve large l1-regularised problems with two-way interactions for binary design
matrices. The novelty of WHInter stems from a new bound to efficiently identify
working sets while avoiding to scan all features, and on fast computations
inspired from solutions to the maximum inner product search problem. We apply
WHInter to simulated and real genetic data and show that it is more scalable
and two orders of magnitude faster than the state of the art.
|
Tensor-based Nonlinear Classifier for High-Order Data Analysis | In this paper we propose a tensor-based nonlinear model for high-order data
classification. The advantages of the proposed scheme are that (i) it
significantly reduces the number of weight parameters, and hence of required
training samples, and (ii) it retains the spatial structure of the input
samples. The proposed model, called \textit{Rank}-1 FNN, is based on a
modification of a feedforward neural network (FNN), such that its weights
satisfy the {\it rank}-1 canonical decomposition. We also introduce a new
learning algorithm to train the model, and we evaluate the \textit{Rank}-1 FNN
on third-order hyperspectral data. Experimental results and comparisons
indicate that the proposed model outperforms state of the art classification
methods, including deep learning based ones, especially in cases with small
numbers of available training samples.
|
Disentangling by Factorising | We define and address the problem of unsupervised learning of disentangled
representations on data generated from independent factors of variation. We
propose FactorVAE, a method that disentangles by encouraging the distribution
of representations to be factorial and hence independent across the dimensions.
We show that it improves upon $\beta$-VAE by providing a better trade-off
between disentanglement and reconstruction quality. Moreover, we highlight the
problems of a commonly used disentanglement metric and introduce a new metric
that does not suffer from them.
|
Improved GQ-CNN: Deep Learning Model for Planning Robust Grasps | Recent developments in the field of robot grasping have shown great
improvements in the grasp success rates when dealing with unknown objects. In
this work we improve on one of the most promising approaches, the Grasp Quality
Convolutional Neural Network (GQ-CNN) trained on the DexNet 2.0 dataset. We
propose a new architecture for the GQ-CNN and describe practical improvements
that increase the model validation accuracy from 92.2% to 95.8% and from 85.9%
to 88.0% on respectively image-wise and object-wise training and validation
splits.
|
Neural Voice Cloning with a Few Samples | Voice cloning is a highly desired feature for personalized speech interfaces.
Neural network based speech synthesis has been shown to generate high quality
speech for a large number of speakers. In this paper, we introduce a neural
voice cloning system that takes a few audio samples as input. We study two
approaches: speaker adaptation and speaker encoding. Speaker adaptation is
based on fine-tuning a multi-speaker generative model with a few cloning
samples. Speaker encoding is based on training a separate model to directly
infer a new speaker embedding from cloning audios and to be used with a
multi-speaker generative model. In terms of naturalness of the speech and its
similarity to original speaker, both approaches can achieve good performance,
even with very few cloning audios. While speaker adaptation can achieve better
naturalness and similarity, the cloning time or required memory for the speaker
encoding approach is significantly less, making it favorable for low-resource
deployment.
|
WebEye - Automated Collection of Malicious HTTP Traffic | With malware detection techniques increasingly adopting machine learning
approaches, the creation of precise training sets becomes more and more
important. Large data sets of realistic web traffic, correctly classified as
benign or malicious are needed, not only to train classic and deep learning
algorithms, but also to serve as evaluation benchmarks for existing malware
detection products. Interestingly, despite the vast number and versatility of
threats a user may encounter when browsing the web, actual malicious content is
often hard to come by, since prerequisites such as browser and operating system
type and version must be met in order to receive the payload from a malware
distributing server. In combination with privacy constraints on data sets of
actual user traffic, it is difficult for researchers and product developers to
evaluate anti-malware solutions against large-scale data sets of realistic web
traffic. In this paper we present WebEye, a framework that autonomously creates
realistic HTTP traffic, enriches recorded traffic with additional information,
and classifies records as malicious or benign, using different classifiers. We
are using WebEye to collect malicious HTML and JavaScript and show how datasets
created with WebEye can be used to train machine learning based malware
detection algorithms. We regard WebEye and the data sets it creates as a tool
for researchers and product developers to evaluate and improve their AI-based
anti-malware solutions against large-scale benchmarks.
|
Orthogonality-Promoting Distance Metric Learning: Convex Relaxation and
Theoretical Analysis | Distance metric learning (DML), which learns a distance metric from labeled
"similar" and "dissimilar" data pairs, is widely utilized. Recently, several
works investigate orthogonality-promoting regularization (OPR), which
encourages the projection vectors in DML to be close to being orthogonal, to
achieve three effects: (1) high balancedness -- achieving comparable
performance on both frequent and infrequent classes; (2) high compactness --
using a small number of projection vectors to achieve a "good" metric; (3) good
generalizability -- alleviating overfitting to training data. While showing
promising results, these approaches suffer three problems. First, they involve
solving non-convex optimization problems where achieving the global optimal is
NP-hard. Second, it lacks a theoretical understanding why OPR can lead to
balancedness. Third, the current generalization error analysis of OPR is not
directly on the regularizer. In this paper, we address these three issues by
(1) seeking convex relaxations of the original nonconvex problems so that the
global optimal is guaranteed to be achievable; (2) providing a formal analysis
on OPR's capability of promoting balancedness; (3) providing a theoretical
analysis that directly reveals the relationship between OPR and generalization
performance. Experiments on various datasets demonstrate that our convex
methods are more effective in promoting balancedness, compactness, and
generalization, and are computationally more efficient, compared with the
nonconvex methods.
|
Policy Evaluation and Optimization with Continuous Treatments | We study the problem of policy evaluation and learning from batched
contextual bandit data when treatments are continuous, going beyond previous
work on discrete treatments. Previous work for discrete treatment/action spaces
focuses on inverse probability weighting (IPW) and doubly robust (DR) methods
that use a rejection sampling approach for evaluation and the equivalent
weighted classification problem for learning. In the continuous setting, this
reduction fails as we would almost surely reject all observations. To tackle
the case of continuous treatments, we extend the IPW and DR approaches to the
continuous setting using a kernel function that leverages treatment proximity
to attenuate discrete rejection. Our policy estimator is consistent and we
characterize the optimal bandwidth. The resulting continuous policy optimizer
(CPO) approach using our estimator achieves convergent regret and approaches
the best-in-class policy for learnable policy classes. We demonstrate that the
estimator performs well and, in particular, outperforms a discretization-based
benchmark. We further study the performance of our policy optimizer in a case
study on personalized dosing based on a dataset of Warfarin patients, their
covariates, and final therapeutic doses. Our learned policy outperforms
benchmarks and nears the oracle-best linear policy.
|
Online Continuous Submodular Maximization | In this paper, we consider an online optimization process, where the
objective functions are not convex (nor concave) but instead belong to a broad
class of continuous submodular functions. We first propose a variant of the
Frank-Wolfe algorithm that has access to the full gradient of the objective
functions. We show that it achieves a regret bound of $O(\sqrt{T})$ (where $T$
is the horizon of the online optimization problem) against a
$(1-1/e)$-approximation to the best feasible solution in hindsight. However, in
many scenarios, only an unbiased estimate of the gradients are available. For
such settings, we then propose an online stochastic gradient ascent algorithm
that also achieves a regret bound of $O(\sqrt{T})$ regret, albeit against a
weaker $1/2$-approximation to the best feasible solution in hindsight. We also
generalize our results to $\gamma$-weakly submodular functions and prove the
same sublinear regret bounds. Finally, we demonstrate the efficiency of our
algorithms on a few problem instances, including non-convex/non-concave
quadratic programs, multilinear extensions of submodular set functions, and
D-optimal design.
|
Variance-based Gradient Compression for Efficient Distributed Deep
Learning | Due to the substantial computational cost, training state-of-the-art deep
neural networks for large-scale datasets often requires distributed training
using multiple computation workers. However, by nature, workers need to
frequently communicate gradients, causing severe bottlenecks, especially on
lower bandwidth connections. A few methods have been proposed to compress
gradient for efficient communication, but they either suffer a low compression
ratio or significantly harm the resulting model accuracy, particularly when
applied to convolutional neural networks. To address these issues, we propose a
method to reduce the communication overhead of distributed deep learning. Our
key observation is that gradient updates can be delayed until an unambiguous
(high amplitude, low variance) gradient has been calculated. We also present an
efficient algorithm to compute the variance with negligible additional cost. We
experimentally show that our method can achieve very high compression ratio
while maintaining the result model accuracy. We also analyze the efficiency
using computation and communication cost models and provide the evidence that
this method enables distributed deep learning for many scenarios with commodity
environments.
|
Bridging Cognitive Programs and Machine Learning | While great advances are made in pattern recognition and machine learning,
the successes of such fields remain restricted to narrow applications and seem
to break down when training data is scarce, a shift in domain occurs, or when
intelligent reasoning is required for rapid adaptation to new environments. In
this work, we list several of the shortcomings of modern machine-learning
solutions, specifically in the contexts of computer vision and in reinforcement
learning and suggest directions to explore in order to try to ameliorate these
weaknesses.
|
Gradient descent with identity initialization efficiently learns
positive definite linear transformations by deep residual networks | We analyze algorithms for approximating a function $f(x) = \Phi x$ mapping
$\Re^d$ to $\Re^d$ using deep linear neural networks, i.e. that learn a
function $h$ parameterized by matrices $\Theta_1,...,\Theta_L$ and defined by
$h(x) = \Theta_L \Theta_{L-1} ... \Theta_1 x$. We focus on algorithms that
learn through gradient descent on the population quadratic loss in the case
that the distribution over the inputs is isotropic.
We provide polynomial bounds on the number of iterations for gradient descent
to approximate the least squares matrix $\Phi$, in the case where the initial
hypothesis $\Theta_1 = ... = \Theta_L = I$ has excess loss bounded by a small
enough constant. On the other hand, we show that gradient descent fails to
converge for $\Phi$ whose distance from the identity is a larger constant, and
we show that some forms of regularization toward the identity in each layer do
not help.
If $\Phi$ is symmetric positive definite, we show that an algorithm that
initializes $\Theta_i = I$ learns an $\epsilon$-approximation of $f$ using a
number of updates polynomial in $L$, the condition number of $\Phi$, and
$\log(d/\epsilon)$. In contrast, we show that if the least squares matrix
$\Phi$ is symmetric and has a negative eigenvalue, then all members of a class
of algorithms that perform gradient descent with identity initialization, and
optionally regularize toward the identity in each layer, fail to converge.
We analyze an algorithm for the case that $\Phi$ satisfies $u^{\top} \Phi u >
0$ for all $u$, but may not be symmetric. This algorithm uses two regularizers:
one that maintains the invariant $u^{\top} \Theta_L \Theta_{L-1} ... \Theta_1 u
> 0$ for all $u$, and another that "balances" $\Theta_1, ..., \Theta_L$ so that
they have the same singular values.
|
Mining Sub-Interval Relationships In Time Series Data | Time-series data is being increasingly collected and stud- ied in several
areas such as neuroscience, climate science, transportation, and social media.
Discovery of complex patterns of relationships between individual time-series,
using data-driven approaches can improve our understanding of real-world
systems. While traditional approaches typically study relationships between two
entire time series, many interesting relationships in real-world applications
exist in small sub-intervals of time while remaining absent or feeble during
other sub-intervals. In this paper, we define the notion of a sub-interval
relationship (SIR) to capture inter- actions between two time series that are
prominent only in certain sub-intervals of time. We propose a novel and
efficient approach to find most interesting SIR in a pair of time series. We
evaluate our proposed approach on two real-world datasets from climate science
and neuroscience domain and demonstrated the scalability and computational
efficiency of our proposed approach. We further evaluated our discovered SIRs
based on a randomization based procedure. Our results indicated the existence
of several such relationships that are statistically significant, some of which
were also found to have physical interpretation.
|
Information-theoretic Limits for Community Detection in Network Models | We analyze the information-theoretic limits for the recovery of node labels
in several network models. This includes the Stochastic Block Model, the
Exponential Random Graph Model, the Latent Space Model, the Directed
Preferential Attachment Model, and the Directed Small-world Model. For the
Stochastic Block Model, the non-recoverability condition depends on the
probabilities of having edges inside a community, and between different
communities. For the Latent Space Model, the non-recoverability condition
depends on the dimension of the latent space, and how far and spread are the
communities in the latent space. For the Directed Preferential Attachment Model
and the Directed Small-world Model, the non-recoverability condition depends on
the ratio between homophily and neighborhood size. We also consider dynamic
versions of the Stochastic Block Model and the Latent Space Model.
|
The Mean-Field Approximation: Information Inequalities, Algorithms, and
Complexity | The mean field approximation to the Ising model is a canonical variational
tool that is used for analysis and inference in Ising models. We provide a
simple and optimal bound for the KL error of the mean field approximation for
Ising models on general graphs, and extend it to higher order Markov random
fields. Our bound improves on previous bounds obtained in work in the graph
limit literature by Borgs, Chayes, Lov\'asz, S\'os, and Vesztergombi and
another recent work by Basak and Mukherjee. Our bound is tight up to lower
order terms. Building on the methods used to prove the bound, along with
techniques from combinatorics and optimization, we study the algorithmic
problem of estimating the (variational) free energy for Ising models and
general Markov random fields. For a graph $G$ on $n$ vertices and interaction
matrix $J$ with Frobenius norm $\| J \|_F$, we provide algorithms that
approximate the free energy within an additive error of $\epsilon n \|J\|_F$ in
time $\exp(poly(1/\epsilon))$. We also show that approximation within $(n
\|J\|_F)^{1-\delta}$ is NP-hard for every $\delta > 0$. Finally, we provide
more efficient approximation algorithms, which find the optimal mean field
approximation, for ferromagnetic Ising models and for Ising models satisfying
Dobrushin's condition.
|
The Vertex Sample Complexity of Free Energy is Polynomial | We study the following question: given a massive Markov random field on $n$
nodes, can a small sample from it provide a rough approximation to the free
energy $\mathcal{F}_n = \log{Z_n}$?
Results in graph limit literature by Borgs, Chayes, Lov\'asz, S\'os, and
Vesztergombi show that for Ising models on $n$ nodes and interactions of
strength $\Theta(1/n)$, an $\epsilon$ approximation to $\log Z_n / n$ can be
achieved by sampling a randomly induced model on $2^{O(1/\epsilon^2)}$ nodes.
We show that the sampling complexity of this problem is {\em polynomial in}
$1/\epsilon$. We further show a polynomial dependence on $\epsilon$ cannot be
avoided.
Our results are very general as they apply to higher order Markov random
fields. For Markov random fields of order $r$, we obtain an algorithm that
achieves $\epsilon$ approximation using a number of samples polynomial in $r$
and $1/\epsilon$ and running time that is $2^{O(1/\epsilon^2)}$ up to
polynomial factors in $r$ and $\epsilon$. For ferromagnetic Ising models, the
running time is polynomial in $1/\epsilon$.
Our results are intimately connected to recent research on the regularity
lemma and property testing, where the interest is in finding which properties
can tested within $\epsilon$ error in time polynomial in $1/\epsilon$. In
particular, our proofs build on results from a recent work by Alon, de la Vega,
Kannan and Karpinski, who also introduced the notion of polynomial vertex
sample complexity. Another critical ingredient of the proof is an effective
bound by the authors of the paper relating the variational free energy and the
free energy.
|
Interaction Matters: A Note on Non-asymptotic Local Convergence of
Generative Adversarial Networks | Motivated by the pursuit of a systematic computational and algorithmic
understanding of Generative Adversarial Networks (GANs), we present a simple
yet unified non-asymptotic local convergence theory for smooth two-player
games, which subsumes several discrete-time gradient-based saddle point
dynamics. The analysis reveals the surprising nature of the off-diagonal
interaction term as both a blessing and a curse. On the one hand, this
interaction term explains the origin of the slow-down effect in the convergence
of Simultaneous Gradient Ascent (SGA) to stable Nash equilibria. On the other
hand, for the unstable equilibria, exponential convergence can be proved thanks
to the interaction term, for four modified dynamics proposed to stabilize GAN
training: Optimistic Mirror Descent (OMD), Consensus Optimization (CO),
Implicit Updates (IU) and Predictive Method (PM). The analysis uncovers the
intimate connections among these stabilizing techniques, and provides detailed
characterization on the choice of learning rate. As a by-product, we present a
new analysis for OMD proposed in Daskalakis, Ilyas, Syrgkanis, and Zeng [2017]
with improved rates.
|
Detecting Social Influence in Event Cascades by Comparing Discriminative
Rankers | The global dynamics of event cascades are often governed by the local
dynamics of peer influence. However, detecting social influence from
observational data is challenging due to confounds like homophily and practical
issues like missing data. We propose a simple discriminative method to detect
influence from observational data. The core of the approach is to train a
ranking algorithm to predict the source of the next event in a cascade, and
compare its out-of-sample accuracy against a competitive baseline which lacks
access to features corresponding to social influence. We analyze synthetically
generated data to show that this method correctly identifies influence in the
presence of confounds, and is robust to both missing data and misspecification
--- unlike well-known alternatives. We apply the method to two real-world
datasets: (1) the co-sponsorship of legislation in the U.S. House of
Representatives on a social network of shared campaign donors; (2) rumors about
the Higgs boson discovery on a follower network of $10^5$ Twitter accounts. Our
model identifies the role of social influence in these scenarios and uses it to
make more accurate predictions about the future trajectory of cascades.
|
Reactive Reinforcement Learning in Asynchronous Environments | The relationship between a reinforcement learning (RL) agent and an
asynchronous environment is often ignored. Frequently used models of the
interaction between an agent and its environment, such as Markov Decision
Processes (MDP) or Semi-Markov Decision Processes (SMDP), do not capture the
fact that, in an asynchronous environment, the state of the environment may
change during computation performed by the agent. In an asynchronous
environment, minimizing reaction time---the time it takes for an agent to react
to an observation---also minimizes the time in which the state of the
environment may change following observation. In many environments, the
reaction time of an agent directly impacts task performance by permitting the
environment to transition into either an undesirable terminal state or a state
where performing the chosen action is inappropriate. We propose a class of
reactive reinforcement learning algorithms that address this problem of
asynchronous environments by immediately acting after observing new state
information. We compare a reactive SARSA learning algorithm with the
conventional SARSA learning algorithm on two asynchronous robotic tasks
(emergency stopping and impact prevention), and show that the reactive RL
algorithm reduces the reaction time of the agent by approximately the duration
of the algorithm's learning update. This new class of reactive algorithms may
facilitate safer control and faster decision making without any change to
standard learning guarantees.
|
A Likelihood-Free Inference Framework for Population Genetic Data using
Exchangeable Neural Networks | An explosion of high-throughput DNA sequencing in the past decade has led to
a surge of interest in population-scale inference with whole-genome data.
Recent work in population genetics has centered on designing inference methods
for relatively simple model classes, and few scalable general-purpose inference
techniques exist for more realistic, complex models. To achieve this, two
inferential challenges need to be addressed: (1) population data are
exchangeable, calling for methods that efficiently exploit the symmetries of
the data, and (2) computing likelihoods is intractable as it requires
integrating over a set of correlated, extremely high-dimensional latent
variables. These challenges are traditionally tackled by likelihood-free
methods that use scientific simulators to generate datasets and reduce them to
hand-designed, permutation-invariant summary statistics, often leading to
inaccurate inference. In this work, we develop an exchangeable neural network
that performs summary statistic-free, likelihood-free inference. Our framework
can be applied in a black-box fashion across a variety of simulation-based
tasks, both within and outside biology. We demonstrate the power of our
approach on the recombination hotspot testing problem, outperforming the
state-of-the-art.
|
CapsuleGAN: Generative Adversarial Capsule Network | We present Generative Adversarial Capsule Network (CapsuleGAN), a framework
that uses capsule networks (CapsNets) instead of the standard convolutional
neural networks (CNNs) as discriminators within the generative adversarial
network (GAN) setting, while modeling image data. We provide guidelines for
designing CapsNet discriminators and the updated GAN objective function, which
incorporates the CapsNet margin loss, for training CapsuleGAN models. We show
that CapsuleGAN outperforms convolutional-GAN at modeling image data
distribution on MNIST and CIFAR-10 datasets, evaluated on the generative
adversarial metric and at semi-supervised image classification.
|
An Alternative View: When Does SGD Escape Local Minima? | Stochastic gradient descent (SGD) is widely used in machine learning.
Although being commonly viewed as a fast but not accurate version of gradient
descent (GD), it always finds better solutions than GD for modern neural
networks.
In order to understand this phenomenon, we take an alternative view that SGD
is working on the convolved (thus smoothed) version of the loss function. We
show that, even if the function $f$ has many bad local minima or saddle points,
as long as for every point $x$, the weighted average of the gradients of its
neighborhoods is one point convex with respect to the desired solution $x^*$,
SGD will get close to, and then stay around $x^*$ with constant probability.
More specifically, SGD will not get stuck at "sharp" local minima with small
diameters, as long as the neighborhoods of these regions contain enough
gradient information. The neighborhood size is controlled by step size and
gradient noise.
Our result identifies a set of functions that SGD provably works, which is
much larger than the set of convex functions. Empirically, we observe that the
loss surface of neural networks enjoys nice one point convexity properties
locally, therefore our theorem helps explain why SGD works so well for neural
networks.
|
Learning to Race through Coordinate Descent Bayesian Optimisation | In the automation of many kinds of processes, the observable outcome can
often be described as the combined effect of an entire sequence of actions, or
controls, applied throughout its execution. In these cases, strategies to
optimise control policies for individual stages of the process might not be
applicable, and instead the whole policy might have to be optimised at once. On
the other hand, the cost to evaluate the policy's performance might also be
high, being desirable that a solution can be found with as few interactions as
possible with the real system. We consider the problem of optimising control
policies to allow a robot to complete a given race track within a minimum
amount of time. We assume that the robot has no prior information about the
track or its own dynamical model, just an initial valid driving example.
Localisation is only applied to monitor the robot and to provide an indication
of its position along the track's centre axis. We propose a method for finding
a policy that minimises the time per lap while keeping the vehicle on the track
using a Bayesian optimisation (BO) approach over a reproducing kernel Hilbert
space. We apply an algorithm to search more efficiently over high-dimensional
policy-parameter spaces with BO, by iterating over each dimension individually,
in a sequential coordinate descent-like scheme. Experiments demonstrate the
performance of the algorithm against other methods in a simulated car racing
environment.
|
Semi-supervised multi-task learning for lung cancer diagnosis | Early detection of lung nodules is of great importance in lung cancer
screening. Existing research recognizes the critical role played by CAD systems
in early detection and diagnosis of lung nodules. However, many CAD systems,
which are used as cancer detection tools, produce a lot of false positives (FP)
and require a further FP reduction step. Furthermore, guidelines for early
diagnosis and treatment of lung cancer are consist of different shape and
volume measurements of abnormalities. Segmentation is at the heart of our
understanding of nodules morphology making it a major area of interest within
the field of computer aided diagnosis systems. This study set out to test the
hypothesis that joint learning of false positive (FP) nodule reduction and
nodule segmentation can improve the computer aided diagnosis (CAD) systems'
performance on both tasks. To support this hypothesis we propose a 3D deep
multi-task CNN to tackle these two problems jointly. We tested our system on
LUNA16 dataset and achieved an average dice similarity coefficient (DSC) of 91%
as segmentation accuracy and a score of nearly 92% for FP reduction. As a proof
of our hypothesis, we showed improvements of segmentation and FP reduction
tasks over two baselines. Our results support that joint training of these two
tasks through a multi-task learning approach improves system performance on
both. We also showed that a semi-supervised approach can be used to overcome
the limitation of lack of labeled data for the 3D segmentation task.
|
CREPE: A Convolutional Representation for Pitch Estimation | The task of estimating the fundamental frequency of a monophonic sound
recording, also known as pitch tracking, is fundamental to audio processing
with multiple applications in speech processing and music information
retrieval. To date, the best performing techniques, such as the pYIN algorithm,
are based on a combination of DSP pipelines and heuristics. While such
techniques perform very well on average, there remain many cases in which they
fail to correctly estimate the pitch. In this paper, we propose a data-driven
pitch tracking algorithm, CREPE, which is based on a deep convolutional neural
network that operates directly on the time-domain waveform. We show that the
proposed model produces state-of-the-art results, performing equally or better
than pYIN. Furthermore, we evaluate the model's generalizability in terms of
noise robustness. A pre-trained version of CREPE is made freely available as an
open-source Python module for easy application.
|
Optimal Single Sample Tests for Structured versus Unstructured Network
Data | We study the problem of testing, using only a single sample, between mean
field distributions (like Curie-Weiss, Erd\H{o}s-R\'enyi) and structured Gibbs
distributions (like Ising model on sparse graphs and Exponential Random
Graphs). Our goal is to test without knowing the parameter values of the
underlying models: only the \emph{structure} of dependencies is known. We
develop a new approach that applies to both the Ising and Exponential Random
Graph settings based on a general and natural statistical test. The test can
distinguish the hypotheses with high probability above a certain threshold in
the (inverse) temperature parameter, and is optimal in that below the threshold
no test can distinguish the hypotheses.
The thresholds do not correspond to the presence of long-range order in the
models. By aggregating information at a global scale, our test works even at
very high temperatures.
The proofs are based on distributional approximation and sharp concentration
of quadratic forms, when restricted to Hamming spheres. The restriction to
Hamming spheres is necessary, since otherwise any scalar statistic is useless
without explicit knowledge of the temperature parameter. At the same time, this
restriction radically changes the behavior of the functions under
consideration, resulting in a much smaller variance than in the independent
setting; this makes it hard to directly apply standard methods (i.e., Stein's
method) for concentration of weakly dependent variables. Instead, we carry out
an additional tensorization argument using a Markov chain that respects the
symmetry of the Hamming sphere.
|
Efficient GAN-Based Anomaly Detection | Generative adversarial networks (GANs) are able to model the complex
highdimensional distributions of real-world data, which suggests they could be
effective for anomaly detection. However, few works have explored the use of
GANs for the anomaly detection task. We leverage recently developed GAN models
for anomaly detection, and achieve state-of-the-art performance on image and
network intrusion datasets, while being several hundred-fold faster at test
time than the only published GAN-based method.
|
A Deep Q-Learning Agent for the L-Game with Variable Batch Training | We employ the Deep Q-Learning algorithm with Experience Replay to train an
agent capable of achieving a high-level of play in the L-Game while
self-learning from low-dimensional states. We also employ variable batch size
for training in order to mitigate the loss of the rare reward signal and
significantly accelerate training. Despite the large action space due to the
number of possible moves, the low-dimensional state space and the rarity of
rewards, which only come at the end of a game, DQL is successful in training an
agent capable of strong play without the use of any search methods or domain
knowledge.
|
A Collaborative Computer Aided Diagnosis (C-CAD) System with
Eye-Tracking, Sparse Attentional Model, and Deep Learning | There are at least two categories of errors in radiology screening that can
lead to suboptimal diagnostic decisions and interventions:(i)human fallibility
and (ii)complexity of visual search. Computer aided diagnostic (CAD) tools are
developed to help radiologists to compensate for some of these errors. However,
despite their significant improvements over conventional screening strategies,
most CAD systems do not go beyond their use as second opinion tools due to
producing a high number of false positives, which human interpreters need to
correct. In parallel with efforts in computerized analysis of radiology scans,
several researchers have examined behaviors of radiologists while screening
medical images to better understand how and why they miss tumors, how they
interact with the information in an image, and how they search for unknown
pathology in the images. Eye-tracking tools have been instrumental in exploring
answers to these fundamental questions. In this paper, we aim to develop a
paradigm shift CAD system, called collaborative CAD (C-CAD), that unifies both
of the above mentioned research lines: CAD and eye-tracking. We design an
eye-tracking interface providing radiologists with a real radiology reading
room experience. Then, we propose a novel algorithm that unifies eye-tracking
data and a CAD system. Specifically, we present a new graph based clustering
and sparsification algorithm to transform eye-tracking data (gaze) into a
signal model to interpret gaze patterns quantitatively and qualitatively. The
proposed C-CAD collaborates with radiologists via eye-tracking technology and
helps them to improve diagnostic decisions. The C-CAD learns radiologists'
search efficiency by processing their gaze patterns. To do this, the C-CAD uses
a deep learning algorithm in a newly designed multi-task learning platform to
segment and diagnose cancers simultaneously.
|
An analysis of training and generalization errors in shallow and deep
networks | This paper is motivated by an open problem around deep networks, namely, the
apparent absence of over-fitting despite large over-parametrization which
allows perfect fitting of the training data. In this paper, we analyze this
phenomenon in the case of regression problems when each unit evaluates a
periodic activation function. We argue that the minimal expected value of the
square loss is inappropriate to measure the generalization error in
approximation of compositional functions in order to take full advantage of the
compositional structure. Instead, we measure the generalization error in the
sense of maximum loss, and sometimes, as a pointwise error. We give estimates
on exactly how many parameters ensure both zero training error as well as a
good generalization error. We prove that a solution of a regularization problem
is guaranteed to yield a good training error as well as a good generalization
error and estimate how much error to expect at which test data.
|
Nonconvex Matrix Factorization from Rank-One Measurements | We consider the problem of recovering low-rank matrices from random rank-one
measurements, which spans numerous applications including covariance sketching,
phase retrieval, quantum state tomography, and learning shallow polynomial
neural networks, among others. Our approach is to directly estimate the
low-rank factor by minimizing a nonconvex quadratic loss function via vanilla
gradient descent, following a tailored spectral initialization. When the true
rank is small, this algorithm is guaranteed to converge to the ground truth (up
to global ambiguity) with near-optimal sample complexity and computational
complexity. To the best of our knowledge, this is the first guarantee that
achieves near-optimality in both metrics. In particular, the key enabler of
near-optimal computational guarantees is an implicit regularization phenomenon:
without explicit regularization, both spectral initialization and the gradient
descent iterates automatically stay within a region incoherent with the
measurement vectors. This feature allows one to employ much more aggressive
step sizes compared with the ones suggested in prior literature, without the
need of sample splitting.
|
Unsupervised vehicle recognition using incremental reseeding of acoustic
signatures | Vehicle recognition and classification have broad applications, ranging from
traffic flow management to military target identification. We demonstrate an
unsupervised method for automated identification of moving vehicles from
roadside audio sensors. Using a short-time Fourier transform to decompose audio
signals, we treat the frequency signature in each time window as an individual
data point. We then use a spectral embedding for dimensionality reduction.
Based on the leading eigenvectors, we relate the performance of an incremental
reseeding algorithm to that of spectral clustering. We find that incremental
reseeding accurately identifies individual vehicles using their acoustic
signatures.
|
Nonparametric Estimation of Low Rank Matrix Valued Function | Let $A:[0,1]\rightarrow\mathbb{H}_m$ (the space of Hermitian matrices) be a
matrix valued function which is low rank with entries in H\"{o}lder class
$\Sigma(\beta,L)$. The goal of this paper is to study statistical estimation of
$A$ based on the regression model $\mathbb{E}(Y_j|\tau_j,X_j) = \langle
A(\tau_j), X_j \rangle,$ where $\tau_j$ are i.i.d. uniformly distributed in
$[0,1]$, $X_j$ are i.i.d. matrix completion sampling matrices, $Y_j$ are
independent bounded responses. We propose an innovative nuclear norm penalized
local polynomial estimator and establish an upper bound on its point-wise risk
measured by Frobenius norm. Then we extend this estimator globally and prove an
upper bound on its integrated risk measured by $L_2$-norm. We also propose
another new estimator based on bias-reducing kernels to study the case when $A$
is not necessarily low rank and establish an upper bound on its risk measured
by $L_{\infty}$-norm. We show that the obtained rates are all optimal up to
some logarithmic factor in minimax sense. Finally, we propose an adaptive
estimation procedure based on Lepskii's method and model selection with data
splitting which is computationally efficient and can be easily implemented and
parallelized.
|
Black-Box Reductions for Parameter-free Online Learning in Banach Spaces | We introduce several new black-box reductions that significantly improve the
design of adaptive and parameter-free online learning algorithms by simplifying
analysis, improving regret guarantees, and sometimes even improving runtime. We
reduce parameter-free online learning to online exp-concave optimization, we
reduce optimization in a Banach space to one-dimensional optimization, and we
reduce optimization over a constrained domain to unconstrained optimization.
All of our reductions run as fast as online gradient descent. We use our new
techniques to improve upon the previously best regret bounds for parameter-free
learning, and do so for arbitrary norms.
|
Exact and Robust Conformal Inference Methods for Predictive Machine
Learning With Dependent Data | We extend conformal inference to general settings that allow for time series
data. Our proposal is developed as a randomization method and accounts for
potential serial dependence by including block structures in the permutation
scheme. As a result, the proposed method retains the exact, model-free validity
when the data are i.i.d. or more generally exchangeable, similar to usual
conformal inference methods. When exchangeability fails, as is the case for
common time series data, the proposed approach is approximately valid under
weak assumptions on the conformity score.
|
Machine learning for Internet of Things data analysis: A survey | Rapid developments in hardware, software, and communication technologies have
allowed the emergence of Internet-connected sensory devices that provide
observation and data measurement from the physical world. By 2020, it is
estimated that the total number of Internet-connected devices being used will
be between 25 and 50 billion. As the numbers grow and technologies become more
mature, the volume of data published will increase. Internet-connected devices
technology, referred to as Internet of Things (IoT), continues to extend the
current Internet by providing connectivity and interaction between the physical
and cyber worlds. In addition to increased volume, the IoT generates Big Data
characterized by velocity in terms of time and location dependency, with a
variety of multiple modalities and varying data quality. Intelligent processing
and analysis of this Big Data is the key to developing smart IoT applications.
This article assesses the different machine learning methods that deal with the
challenges in IoT data by considering smart cities as the main use case. The
key contribution of this study is presentation of a taxonomy of machine
learning algorithms explaining how different techniques are applied to the data
in order to extract higher level information. The potential and challenges of
machine learning for IoT data analytics will also be discussed. A use case of
applying Support Vector Machine (SVM) on Aarhus Smart City traffic data is
presented for a more detailed exploration.
|
Learning Data-Driven Objectives to Optimize Interactive Systems | Effective optimization is essential for interactive systems to provide a
satisfactory user experience. However, it is often challenging to find an
objective to optimize for. Generally, such objectives are manually crafted and
rarely capture complex user needs in an accurate manner. We propose an approach
that infers the objective directly from observed user interactions. These
inferences can be made regardless of prior knowledge and across different types
of user behavior. We introduce interactive system optimization, a novel
algorithm that uses these inferred objectives for optimization. Our main
contribution is a new general principled approach to optimizing interactive
systems using data-driven objectives. We demonstrate the high effectiveness of
interactive system optimization over several simulations.
|
Learning Adversarially Fair and Transferable Representations | In this paper, we advocate for representation learning as the key to
mitigating unfair prediction outcomes downstream. Motivated by a scenario where
learned representations are used by third parties with unknown objectives, we
propose and explore adversarial representation learning as a natural method of
ensuring those parties act fairly. We connect group fairness (demographic
parity, equalized odds, and equal opportunity) to different adversarial
objectives. Through worst-case theoretical guarantees and experimental
validation, we show that the choice of this objective is crucial to fair
prediction. Furthermore, we present the first in-depth experimental
demonstration of fair transfer learning and demonstrate empirically that our
learned representations admit fair predictions on new tasks while maintaining
utility, an essential goal of fair representation learning.
|
Sequence-to-Sequence Prediction of Vehicle Trajectory via LSTM
Encoder-Decoder Architecture | In this paper, we propose a deep learning based vehicle trajectory prediction
technique which can generate the future trajectory sequence of surrounding
vehicles in real time. We employ the encoder-decoder architecture which
analyzes the pattern underlying in the past trajectory using the long
short-term memory (LSTM) based encoder and generates the future trajectory
sequence using the LSTM based decoder. This structure produces the $K$ most
likely trajectory candidates over occupancy grid map by employing the beam
search technique which keeps the $K$ locally best candidates from the decoder
output. The experiments conducted on highway traffic scenarios show that the
prediction accuracy of the proposed method is significantly higher than the
conventional trajectory prediction techniques.
|
Stochastic Chebyshev Gradient Descent for Spectral Optimization | A large class of machine learning techniques requires the solution of
optimization problems involving spectral functions of parametric matrices, e.g.
log-determinant and nuclear norm. Unfortunately, computing the gradient of a
spectral function is generally of cubic complexity, as such gradient descent
methods are rather expensive for optimizing objectives involving the spectral
function. Thus, one naturally turns to stochastic gradient methods in hope that
they will provide a way to reduce or altogether avoid the computation of full
gradients. However, here a new challenge appears: there is no straightforward
way to compute unbiased stochastic gradients for spectral functions. In this
paper, we develop unbiased stochastic gradients for spectral-sums, an important
subclass of spectral functions. Our unbiased stochastic gradients are based on
combining randomized trace estimators with stochastic truncation of the
Chebyshev expansions. A careful design of the truncation distribution allows us
to offer distributions that are variance-optimal, which is crucial for fast and
stable convergence of stochastic gradient methods. We further leverage our
proposed stochastic gradients to devise stochastic methods for objective
functions involving spectral-sums, and rigorously analyze their convergence
rate. The utility of our methods is demonstrated in numerical experiments.
|
Convergence of Online Mirror Descent | In this paper we consider online mirror descent (OMD) algorithms, a class of
scalable online learning algorithms exploiting data geometric structures
through mirror maps. Necessary and sufficient conditions are presented in terms
of the step size sequence $\{\eta_t\}_{t}$ for the convergence of an OMD
algorithm with respect to the expected Bregman distance induced by the mirror
map. The condition is $\lim_{t\to\infty}\eta_t=0,
\sum_{t=1}^{\infty}\eta_t=\infty$ in the case of positive variances. It is
reduced to $\sum_{t=1}^{\infty}\eta_t=\infty$ in the case of zero variances for
which the linear convergence may be achieved by taking a constant step size
sequence. A sufficient condition on the almost sure convergence is also given.
We establish tight error bounds under mild conditions on the mirror map, the
loss function, and the regularizer. Our results are achieved by some novel
analysis on the one-step progress of the OMD algorithm using smoothness and
strong convexity of the mirror map and the loss function.
|
Anomaly Detection using One-Class Neural Networks | We propose a one-class neural network (OC-NN) model to detect anomalies in
complex data sets. OC-NN combines the ability of deep networks to extract a
progressively rich representation of data with the one-class objective of
creating a tight envelope around normal data. The OC-NN approach breaks new
ground for the following crucial reason: data representation in the hidden
layer is driven by the OC-NN objective and is thus customized for anomaly
detection. This is a departure from other approaches which use a hybrid
approach of learning deep features using an autoencoder and then feeding the
features into a separate anomaly detection method like one-class SVM (OC-SVM).
The hybrid OC-SVM approach is sub-optimal because it is unable to influence
representational learning in the hidden layers. A comprehensive set of
experiments demonstrate that on complex data sets (like CIFAR and GTSRB), OC-NN
performs on par with state-of-the-art methods and outperformed conventional
shallow methods in some scenarios.
|
Efficient Sparse-Winograd Convolutional Neural Networks | Convolutional Neural Networks (CNNs) are computationally intensive, which
limits their application on mobile devices. Their energy is dominated by the
number of multiplies needed to perform the convolutions. Winograd's minimal
filtering algorithm (Lavin, 2015) and network pruning (Han et al., 2015) can
reduce the operation count, but these two methods cannot be directly combined
$-$ applying the Winograd transform fills in the sparsity in both the weights
and the activations. We propose two modifications to Winograd-based CNNs to
enable these methods to exploit sparsity. First, we move the ReLU operation
into the Winograd domain to increase the sparsity of the transformed
activations. Second, we prune the weights in the Winograd domain to exploit
static weight sparsity. For models on CIFAR-10, CIFAR-100 and ImageNet
datasets, our method reduces the number of multiplications by $10.4\times$,
$6.8\times$ and $10.8\times$ respectively with loss of accuracy less than
$0.1\%$, outperforming previous baselines by $2.0\times$-$3.0\times$. We also
show that moving ReLU to the Winograd domain allows more aggressive pruning.
|
Node Centralities and Classification Performance for Characterizing Node
Embedding Algorithms | Embedding graph nodes into a vector space can allow the use of machine
learning to e.g. predict node classes, but the study of node embedding
algorithms is immature compared to the natural language processing field
because of a diverse nature of graphs. We examine the performance of node
embedding algorithms with respect to graph centrality measures that
characterize diverse graphs, through systematic experiments with four node
embedding algorithms, four or five graph centralities, and six datasets.
Experimental results give insights into the properties of node embedding
algorithms, which can be a basis for further research on this topic.
|
Inductive Framework for Multi-Aspect Streaming Tensor Completion with
Side Information | Low rank tensor completion is a well studied problem and has applications in
various fields. However, in many real world applications the data is dynamic,
i.e., new data arrives at different time intervals. As a result, the tensors
used to represent the data grow in size. Besides the tensors, in many real
world scenarios, side information is also available in the form of matrices
which also grow in size with time. The problem of predicting missing values in
the dynamically growing tensor is called dynamic tensor completion. Most of the
previous work in dynamic tensor completion make an assumption that the tensor
grows only in one mode. To the best of our Knowledge, there is no previous work
which incorporates side information with dynamic tensor completion. We bridge
this gap in this paper by proposing a dynamic tensor completion framework
called Side Information infused Incremental Tensor Analysis (SIITA), which
incorporates side information and works for general incremental tensors. We
also show how non-negative constraints can be incorporated with SIITA, which is
essential for mining interpretable latent clusters. We carry out extensive
experiments on multiple real world datasets to demonstrate the effectiveness of
SIITA in various different settings.
|
Space-efficient Feature Maps for String Alignment Kernels | String kernels are attractive data analysis tools for analyzing string data.
Among them, alignment kernels are known for their high prediction accuracies in
string classifications when tested in combination with SVM in various
applications. However, alignment kernels have a crucial drawback in that they
scale poorly due to their quadratic computation complexity in the number of
input strings, which limits large-scale applications in practice. We address
this need by presenting the first approximation for string alignment kernels,
which we call space-efficient feature maps for edit distance with moves
(SFMEDM), by leveraging a metric embedding named edit sensitive parsing (ESP)
and feature maps (FMs) of random Fourier features (RFFs) for large-scale string
analyses. The original FMs for RFFs consume a huge amount of memory
proportional to the dimension d of input vectors and the dimension D of output
vectors, which prohibits its large-scale applications. We present novel
space-efficient feature maps (SFMs) of RFFs for a space reduction from O(dD) of
the original FMs to O(d) of SFMs with a theoretical guarantee with respect to
concentration bounds. We experimentally test SFMEDM on its ability to learn SVM
for large-scale string classifications with various massive string data, and we
demonstrate the superior performance of SFMEDM with respect to prediction
accuracy, scalability and computation efficiency.
|
Efficient Gaussian Process Classification Using Polya-Gamma Data
Augmentation | We propose a scalable stochastic variational approach to GP classification
building on Polya-Gamma data augmentation and inducing points. Unlike former
approaches, we obtain closed-form updates based on natural gradients that lead
to efficient optimization. We evaluate the algorithm on real-world datasets
containing up to 11 million data points and demonstrate that it is up to two
orders of magnitude faster than the state-of-the-art while being competitive in
terms of prediction performance.
|
Spurious Valleys in Two-layer Neural Network Optimization Landscapes | Neural networks provide a rich class of high-dimensional, non-convex
optimization problems. Despite their non-convexity, gradient-descent methods
often successfully optimize these models. This has motivated a recent spur in
research attempting to characterize properties of their loss surface that may
explain such success.
In this paper, we address this phenomenon by studying a key topological
property of the loss: the presence or absence of spurious valleys, defined as
connected components of sub-level sets that do not include a global minimum.
Focusing on a class of two-layer neural networks defined by smooth (but
generally non-linear) activation functions, we identify a notion of intrinsic
dimension and show that it provides necessary and sufficient conditions for the
absence of spurious valleys. More concretely, finite intrinsic dimension
guarantees that for sufficiently overparametrised models no spurious valleys
exist, independently of the data distribution. Conversely, infinite intrinsic
dimension implies that spurious valleys do exist for certain data
distributions, independently of model overparametrisation. Besides these
positive and negative results, we show that, although spurious valleys may
exist in general, they are confined to low risk levels and avoided with high
probability on overparametrised models.
|
Training Big Random Forests with Little Resources | Without access to large compute clusters, building random forests on large
datasets is still a challenging problem. This is, in particular, the case if
fully-grown trees are desired. We propose a simple yet effective framework that
allows to efficiently construct ensembles of huge trees for hundreds of
millions or even billions of training instances using a cheap desktop computer
with commodity hardware. The basic idea is to consider a multi-level
construction scheme, which builds top trees for small random subsets of the
available data and which subsequently distributes all training instances to the
top trees' leaves for further processing. While being conceptually simple, the
overall efficiency crucially depends on the particular implementation of the
different phases. The practical merits of our approach are demonstrated using
dense datasets with hundreds of millions of training instances.
|
HybridSVD: When Collaborative Information is Not Enough | We propose a new hybrid algorithm that allows incorporating both user and
item side information within the standard collaborative filtering technique.
One of its key features is that it naturally extends a simple PureSVD approach
and inherits its unique advantages, such as highly efficient Lanczos-based
optimization procedure, simplified hyper-parameter tuning and a quick
folding-in computation for generating recommendations instantly even in highly
dynamic online environments. The algorithm utilizes a generalized formulation
of the singular value decomposition, which adds flexibility to the solution and
allows imposing the desired structure on its latent space. Conveniently, the
resulting model also admits an efficient and straightforward solution for the
cold start scenario. We evaluate our approach on a diverse set of datasets and
show its superiority over similar classes of hybrid models.
|
Towards Ultra-High Performance and Energy Efficiency of Deep Learning
Systems: An Algorithm-Hardware Co-Optimization Framework | Hardware accelerations of deep learning systems have been extensively
investigated in industry and academia. The aim of this paper is to achieve
ultra-high energy efficiency and performance for hardware implementations of
deep neural networks (DNNs). An algorithm-hardware co-optimization framework is
developed, which is applicable to different DNN types, sizes, and application
scenarios. The algorithm part adopts the general block-circulant matrices to
achieve a fine-grained tradeoff between accuracy and compression ratio. It
applies to both fully-connected and convolutional layers and contains a
mathematically rigorous proof of the effectiveness of the method. The proposed
algorithm reduces computational complexity per layer from O($n^2$) to O($n\log
n$) and storage complexity from O($n^2$) to O($n$), both for training and
inference. The hardware part consists of highly efficient Field Programmable
Gate Array (FPGA)-based implementations using effective reconfiguration, batch
processing, deep pipelining, resource re-using, and hierarchical control.
Experimental results demonstrate that the proposed framework achieves at least
152X speedup and 71X energy efficiency gain compared with IBM TrueNorth
processor under the same test accuracy. It achieves at least 31X energy
efficiency gain compared with the reference FPGA-based work.
|
RadialGAN: Leveraging multiple datasets to improve target-specific
predictive models using Generative Adversarial Networks | Training complex machine learning models for prediction often requires a
large amount of data that is not always readily available. Leveraging these
external datasets from related but different sources is therefore an important
task if good predictive models are to be built for deployment in settings where
data can be rare. In this paper we propose a novel approach to the problem in
which we use multiple GAN architectures to learn to translate from one dataset
to another, thereby allowing us to effectively enlarge the target dataset, and
therefore learn better predictive models than if we simply used the target
dataset. We show the utility of such an approach, demonstrating that our method
improves the prediction performance on the target domain over using just the
target dataset and also show that our framework outperforms several other
benchmarks on a collection of real-world medical datasets.
|
Sim-to-Real Optimization of Complex Real World Mobile Network with
Imperfect Information via Deep Reinforcement Learning from Self-play | Mobile network that millions of people use every day is one of the most
complex systems in the world. Optimization of mobile network to meet exploding
customer demand and reduce capital/operation expenditures poses great
challenges. Despite recent progress, application of deep reinforcement learning
(DRL) to complex real world problem still remains unsolved, given data
scarcity, partial observability, risk and complex rules/dynamics in real world,
as well as the huge reality gap between simulation and real world. To bridge
the reality gap, we introduce a Sim-to-Real framework to directly transfer
learning from simulation to real world via graph convolutional neural network
(CNN) - by abstracting partially observable mobile network into graph, then
distilling domain-variant irregular graph into domain-invariant tensor in
locally Euclidean space as input to CNN -, domain randomization and multi-task
learning. We use a novel self-play mechanism to encourage competition among DRL
agents for best record on multiple tasks via simulated annealing, just like
athletes compete for world record in decathlon. We also propose a decentralized
multi-agent, competitive and cooperative DRL method to coordinate the actions
of multi-cells to maximize global reward and minimize negative impact to
neighbor cells. Using 6 field trials on commercial mobile networks, we
demonstrate for the first time that a DRL agent can successfully transfer
learning from simulation to complex real world problem with imperfect
information, complex rules/dynamics, huge state/action space, and multi-agent
interactions, without any training in the real world.
|
Improving Mild Cognitive Impairment Prediction via Reinforcement
Learning and Dialogue Simulation | Mild cognitive impairment (MCI) is a prodromal phase in the progression from
normal aging to dementia, especially Alzheimers disease. Even though there is
mild cognitive decline in MCI patients, they have normal overall cognition and
thus is challenging to distinguish from normal aging. Using transcribed data
obtained from recorded conversational interactions between participants and
trained interviewers, and applying supervised learning models to these data, a
recent clinical trial has shown a promising result in differentiating MCI from
normal aging. However, the substantial amount of interactions with medical
staff can still incur significant medical care expenses in practice. In this
paper, we propose a novel reinforcement learning (RL) framework to train an
efficient dialogue agent on existing transcripts from clinical trials.
Specifically, the agent is trained to sketch disease-specific lexical
probability distribution, and thus to converse in a way that maximizes the
diagnosis accuracy and minimizes the number of conversation turns. We evaluate
the performance of the proposed reinforcement learning framework on the MCI
diagnosis from a real clinical trial. The results show that while using only a
few turns of conversation, our framework can significantly outperform
state-of-the-art supervised learning approaches.
|
Music Genre Classification using Masked Conditional Neural Networks | The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural
Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The
CLNN captures the conditional temporal influence between the frames in a window
and the mask in the MCLNN enforces a systematic sparseness that follows a
filterbank-like pattern over the network links. The mask induces the network to
learn about time-frequency representations in bands, allowing the network to
sustain frequency shifts. Additionally, the mask in the MCLNN automates the
exploration of a range of feature combinations, usually done through an
exhaustive manual search. We have evaluated the MCLNN performance using the
Ballroom and Homburg datasets of music genres. MCLNN has achieved accuracies
that are competitive to state-of-the-art handcrafted attempts in addition to
models based on Convolutional Neural Networks.
|
Local Optimality and Generalization Guarantees for the Langevin
Algorithm via Empirical Metastability | We study the detailed path-wise behavior of the discrete-time Langevin
algorithm for non-convex Empirical Risk Minimization (ERM) through the lens of
metastability, adopting some techniques from Berglund and Gentz (2003.
For a particular local optimum of the empirical risk, with an arbitrary
initialization, we show that, with high probability, at least one of the
following two events will occur: (1) the Langevin trajectory ends up somewhere
outside the $\varepsilon$-neighborhood of this particular optimum within a
short recurrence time; (2) it enters this $\varepsilon$-neighborhood by the
recurrence time and stays there until a potentially exponentially long escape
time. We call this phenomenon empirical metastability.
This two-timescale characterization aligns nicely with the existing
literature in the following two senses. First, the effective recurrence time
(i.e., number of iterations multiplied by stepsize) is dimension-independent,
and resembles the convergence time of continuous-time deterministic Gradient
Descent (GD). However unlike GD, the Langevin algorithm does not require strong
conditions on local initialization, and has the possibility of eventually
visiting all optima. Second, the scaling of the escape time is consistent with
the Eyring-Kramers law, which states that the Langevin scheme will eventually
visit all local minima, but it will take an exponentially long time to transit
among them. We apply this path-wise concentration result in the context of
statistical learning to examine local notions of generalization and optimality.
|
A Generative Modeling Approach to Limited Channel ECG Classification | Processing temporal sequences is central to a variety of applications in
health care, and in particular multi-channel Electrocardiogram (ECG) is a
highly prevalent diagnostic modality that relies on robust sequence modeling.
While Recurrent Neural Networks (RNNs) have led to significant advances in
automated diagnosis with time-series data, they perform poorly when models are
trained using a limited set of channels. A crucial limitation of existing
solutions is that they rely solely on discriminative models, which tend to
generalize poorly in such scenarios. In order to combat this limitation, we
develop a generative modeling approach to limited channel ECG classification.
This approach first uses a Seq2Seq model to implicitly generate the missing
channel information, and then uses the latent representation to perform the
actual supervisory task. This decoupling enables the use of unsupervised data
and also provides highly robust metric spaces for subsequent discriminative
learning. Our experiments with the Physionet dataset clearly evidence the
effectiveness of our approach over standard RNNs in disease prediction.
|
Guaranteed Recovery of One-Hidden-Layer Neural Networks via Cross
Entropy | We study model recovery for data classification, where the training labels
are generated from a one-hidden-layer neural network with sigmoid activations,
also known as a single-layer feedforward network, and the goal is to recover
the weights of the neural network. We consider two network models, the
fully-connected network (FCN) and the non-overlapping convolutional neural
network (CNN). We prove that with Gaussian inputs, the empirical risk based on
cross entropy exhibits strong convexity and smoothness {\em uniformly} in a
local neighborhood of the ground truth, as soon as the sample complexity is
sufficiently large. This implies that if initialized in this neighborhood,
gradient descent converges linearly to a critical point that is provably close
to the ground truth. Furthermore, we show such an initialization can be
obtained via the tensor method. This establishes the global convergence
guarantee for empirical risk minimization using cross entropy via gradient
descent for learning one-hidden-layer neural networks, at the near-optimal
sample and computational complexity with respect to the network input dimension
without unrealistic assumptions such as requiring a fresh set of samples at
each iteration.
|
Recurrent Binary Embedding for GPU-Enabled Exhaustive Retrieval from
Billion-Scale Semantic Vectors | Rapid advances in GPU hardware and multiple areas of Deep Learning open up a
new opportunity for billion-scale information retrieval with exhaustive search.
Building on top of the powerful concept of semantic learning, this paper
proposes a Recurrent Binary Embedding (RBE) model that learns compact
representations for real-time retrieval. The model has the unique ability to
refine a base binary vector by progressively adding binary residual vectors to
meet the desired accuracy. The refined vector enables efficient implementation
of exhaustive similarity computation with bit-wise operations, followed by a
near- lossless k-NN selection algorithm, also proposed in this paper. The
proposed algorithms are integrated into an end-to-end multi-GPU system that
retrieves thousands of top items from over a billion candidates in real-time.
The RBE model and the retrieval system were evaluated with data from a major
paid search engine. When measured against the state-of-the-art model for binary
representation and the full precision model for semantic embedding, RBE
significantly outperformed the former, and filled in over 80% of the AUC gap
in-between. Experiments comparing with our production retrieval system also
demonstrated superior performance. While the primary focus of this paper is to
build RBE based on a particular class of semantic models, generalizing to other
types is straightforward, as exemplified by two different models at the end of
the paper.
|
Memorize or generalize? Searching for a compositional RNN in a haystack | Neural networks are very powerful learning systems, but they do not readily
generalize from one task to the other. This is partly due to the fact that they
do not learn in a compositional way, that is, by discovering skills that are
shared by different tasks, and recombining them to solve new problems. In this
paper, we explore the compositional generalization capabilities of recurrent
neural networks (RNNs). We first propose the lookup table composition domain as
a simple setup to test compositional behaviour and show that it is
theoretically possible for a standard RNN to learn to behave compositionally in
this domain when trained with standard gradient descent and provided with
additional supervision. We then remove this additional supervision and perform
a search over a large number of model initializations to investigate the
proportion of RNNs that can still converge to a compositional solution. We
discover that a small but non-negligible proportion of RNNs do reach partial
compositional solutions even without special architectural constraints. This
suggests that a combination of gradient descent and evolutionary strategies
directly favouring the minority models that developed more compositional
approaches might suffice to lead standard RNNs towards compositional solutions.
|
Online convex optimization for cumulative constraints | We propose the algorithms for online convex optimization which lead to
cumulative squared constraint violations of the form
$\sum\limits_{t=1}^T\big([g(x_t)]_+\big)^2=O(T^{1-\beta})$, where
$\beta\in(0,1)$. Previous literature has focused on long-term constraints of
the form $\sum\limits_{t=1}^Tg(x_t)$. There, strictly feasible solutions can
cancel out the effects of violated constraints. In contrast, the new form
heavily penalizes large constraint violations and cancellation effects cannot
occur.
Furthermore, useful bounds on the single step constraint violation
$[g(x_t)]_+$ are derived.
For convex objectives, our regret bounds generalize existing bounds, and for
strongly convex objectives we give improved regret bounds.
In numerical experiments, we show that our algorithm closely follows the
constraint boundary leading to low cumulative violation.
|
Simultaneous Modeling of Multiple Complications for Risk Profiling in
Diabetes Care | Type 2 diabetes mellitus (T2DM) is a chronic disease that often results in
multiple complications. Risk prediction and profiling of T2DM complications is
critical for healthcare professionals to design personalized treatment plans
for patients in diabetes care for improved outcomes. In this paper, we study
the risk of developing complications after the initial T2DM diagnosis from
longitudinal patient records. We propose a novel multi-task learning approach
to simultaneously model multiple complications where each task corresponds to
the risk modeling of one complication. Specifically, the proposed method
strategically captures the relationships (1) between the risks of multiple T2DM
complications, (2) between the different risk factors, and (3) between the risk
factor selection patterns. The method uses coefficient shrinkage to identify an
informative subset of risk factors from high-dimensional data, and uses a
hierarchical Bayesian framework to allow domain knowledge to be incorporated as
priors. The proposed method is favorable for healthcare applications because in
additional to improved prediction performance, relationships among the
different risks and risk factors are also identified. Extensive experimental
results on a large electronic medical claims database show that the proposed
method outperforms state-of-the-art models by a significant margin.
Furthermore, we show that the risk associations learned and the risk factors
identified lead to meaningful clinical insights.
|
Accelerated Primal-Dual Policy Optimization for Safe Reinforcement
Learning | Constrained Markov Decision Process (CMDP) is a natural framework for
reinforcement learning tasks with safety constraints, where agents learn a
policy that maximizes the long-term reward while satisfying the constraints on
the long-term cost. A canonical approach for solving CMDPs is the primal-dual
method which updates parameters in primal and dual spaces in turn. Existing
methods for CMDPs only use on-policy data for dual updates, which results in
sample inefficiency and slow convergence. In this paper, we propose a policy
search method for CMDPs called Accelerated Primal-Dual Optimization (APDO),
which incorporates an off-policy trained dual variable in the dual update
procedure while updating the policy in primal space with on-policy likelihood
ratio gradient. Experimental results on a simulated robot locomotion task show
that APDO achieves better sample efficiency and faster convergence than
state-of-the-art approaches for CMDPs.
|
Robust Estimation via Robust Gradient Estimation | We provide a new computationally-efficient class of estimators for risk
minimization. We show that these estimators are robust for general statistical
models: in the classical Huber epsilon-contamination model and in heavy-tailed
settings. Our workhorse is a novel robust variant of gradient descent, and we
provide conditions under which our gradient descent variant provides accurate
estimators in a general convex risk minimization problem. We provide specific
consequences of our theory for linear regression, logistic regression and for
estimation of the canonical parameters in an exponential family. These results
provide some of the first computationally tractable and provably robust
estimators for these canonical statistical models. Finally, we study the
empirical performance of our proposed methods on synthetic and real datasets,
and find that our methods convincingly outperform a variety of baselines.
|
Recommendations with Negative Feedback via Pairwise Deep Reinforcement
Learning | Recommender systems play a crucial role in mitigating the problem of
information overload by suggesting users' personalized items or services. The
vast majority of traditional recommender systems consider the recommendation
procedure as a static process and make recommendations following a fixed
strategy. In this paper, we propose a novel recommender system with the
capability of continuously improving its strategies during the interactions
with users. We model the sequential interactions between users and a
recommender system as a Markov Decision Process (MDP) and leverage
Reinforcement Learning (RL) to automatically learn the optimal strategies via
recommending trial-and-error items and receiving reinforcements of these items
from users' feedback. Users' feedback can be positive and negative and both
types of feedback have great potentials to boost recommendations. However, the
number of negative feedback is much larger than that of positive one; thus
incorporating them simultaneously is challenging since positive feedback could
be buried by negative one. In this paper, we develop a novel approach to
incorporate them into the proposed deep recommender system (DEERS) framework.
The experimental results based on real-world e-commerce data demonstrate the
effectiveness of the proposed framework. Further experiments have been
conducted to understand the importance of both positive and negative feedback
in recommendations.
|
EA-CG: An Approximate Second-Order Method for Training Fully-Connected
Neural Networks | For training fully-connected neural networks (FCNNs), we propose a practical
approximate second-order method including: 1) an approximation of the Hessian
matrix and 2) a conjugate gradient (CG) based method. Our proposed approximate
Hessian matrix is memory-efficient and can be applied to any FCNNs where the
activation and criterion functions are twice differentiable. We devise a
CG-based method incorporating one-rank approximation to derive Newton
directions for training FCNNs, which significantly reduces both space and time
complexity. This CG-based method can be employed to solve any linear equation
where the coefficient matrix is Kronecker-factored, symmetric and positive
definite. Empirical studies show the efficacy and efficiency of our proposed
method.
|
On the Optimization of Deep Networks: Implicit Acceleration by
Overparameterization | Conventional wisdom in deep learning states that increasing depth improves
expressiveness but complicates optimization. This paper suggests that,
sometimes, increasing depth can speed up optimization. The effect of depth on
optimization is decoupled from expressiveness by focusing on settings where
additional layers amount to overparameterization - linear neural networks, a
well-studied model. Theoretical analysis, as well as experiments, show that
here depth acts as a preconditioner which may accelerate convergence. Even on
simple convex problems such as linear regression with $\ell_p$ loss, $p>2$,
gradient descent can benefit from transitioning to a non-convex
overparameterized objective, more than it would from some common acceleration
schemes. We also prove that it is mathematically impossible to obtain the
acceleration effect of overparametrization via gradients of any regularizer.
|
Subspace Network: Deep Multi-Task Censored Regression for Modeling
Neurodegenerative Diseases | Over the past decade a wide spectrum of machine learning models have been
developed to model the neurodegenerative diseases, associating biomarkers,
especially non-intrusive neuroimaging markers, with key clinical scores
measuring the cognitive status of patients. Multi-task learning (MTL) has been
commonly utilized by these studies to address high dimensionality and small
cohort size challenges. However, most existing MTL approaches are based on
linear models and suffer from two major limitations: 1) they cannot explicitly
consider upper/lower bounds in these clinical scores; 2) they lack the
capability to capture complicated non-linear interactions among the variables.
In this paper, we propose Subspace Network, an efficient deep modeling approach
for non-linear multi-task censored regression. Each layer of the subspace
network performs a multi-task censored regression to improve upon the
predictions from the last layer via sketching a low-dimensional subspace to
perform knowledge transfer among learning tasks. Under mild assumptions, for
each layer the parametric subspace can be recovered using only one pass of
training data. Empirical results demonstrate that the proposed subspace network
quickly picks up the correct parameter subspaces, and outperforms
state-of-the-arts in predicting neurodegenerative clinical scores using
information in brain imaging.
|
Heron Inference for Bayesian Graphical Models | Bayesian graphical models have been shown to be a powerful tool for
discovering uncertainty and causal structure from real-world data in many
application fields. Current inference methods primarily follow different kinds
of trade-offs between computational complexity and predictive accuracy. At one
end of the spectrum, variational inference approaches perform well in
computational efficiency, while at the other end, Gibbs sampling approaches are
known to be relatively accurate for prediction in practice. In this paper, we
extend an existing Gibbs sampling method, and propose a new deterministic Heron
inference (Heron) for a family of Bayesian graphical models. In addition to the
support for nontrivial distributability, one more benefit of Heron is that it
is able to not only allow us to easily assess the convergence status but also
largely improve the running efficiency. We evaluate Heron against the standard
collapsed Gibbs sampler and state-of-the-art state augmentation method in
inference for well-known graphical models. Experimental results using publicly
available real-life data have demonstrated that Heron significantly outperforms
the baseline methods for inferring Bayesian graphical models.
|
Are Generative Classifiers More Robust to Adversarial Attacks? | There is a rising interest in studying the robustness of deep neural network
classifiers against adversaries, with both advanced attack and defence
techniques being actively developed. However, most recent work focuses on
discriminative classifiers, which only model the conditional distribution of
the labels given the inputs. In this paper, we propose and investigate the deep
Bayes classifier, which improves classical naive Bayes with conditional deep
generative models. We further develop detection methods for adversarial
examples, which reject inputs with low likelihood under the generative model.
Experimental results suggest that deep Bayes classifiers are more robust than
deep discriminative classifiers, and that the proposed detection methods are
effective against many recently proposed attacks.
|
Robustness of Rotation-Equivariant Networks to Adversarial Perturbations | Deep neural networks have been shown to be vulnerable to adversarial
examples: very small perturbations of the input having a dramatic impact on the
predictions. A wealth of adversarial attacks and distance metrics to quantify
the similarity between natural and adversarial images have been proposed,
recently enlarging the scope of adversarial examples with geometric
transformations beyond pixel-wise attacks. In this context, we investigate the
robustness to adversarial attacks of new Convolutional Neural Network
architectures providing equivariance to rotations. We found that
rotation-equivariant networks are significantly less vulnerable to
geometric-based attacks than regular networks on the MNIST, CIFAR-10, and
ImageNet datasets.
|
Finding Influential Training Samples for Gradient Boosted Decision Trees | We address the problem of finding influential training samples for a
particular case of tree ensemble-based models, e.g., Random Forest (RF) or
Gradient Boosted Decision Trees (GBDT). A natural way of formalizing this
problem is studying how the model's predictions change upon leave-one-out
retraining, leaving out each individual training sample. Recent work has shown
that, for parametric models, this analysis can be conducted in a
computationally efficient way. We propose several ways of extending this
framework to non-parametric GBDT ensembles under the assumption that tree
structures remain fixed. Furthermore, we introduce a general scheme of
obtaining further approximations to our method that balance the trade-off
between performance and computational complexity. We evaluate our approaches on
various experimental setups and use-case scenarios and demonstrate both the
quality of our approach to finding influential training samples in comparison
to the baselines and its computational efficiency.
|
Degeneration in VAE: in the Light of Fisher Information Loss | While enormous progress has been made to Variational Autoencoder (VAE) in
recent years, similar to other deep networks, VAE with deep networks suffers
from the problem of degeneration, which seriously weakens the correlation
between the input and the corresponding latent codes, deviating from the goal
of the representation learning. To investigate how degeneration affects VAE
from a theoretical perspective, we illustrate the information transmission in
VAE and analyze the intermediate layers of the encoders/decoders. Specifically,
we propose a Fisher Information measure for the layer-wise analysis. With such
measure, we demonstrate that information loss is ineluctable in feed-forward
networks and causes the degeneration in VAE. We show that skip connections in
VAE enable the preservation of information without changing the model
architecture. We call this class of VAE equipped with skip connections as SCVAE
and perform a range of experiments to show its advantages in information
preservation and degeneration mitigation.
|
Deep Echo State Networks for Diagnosis of Parkinson's Disease | In this paper, we introduce a novel approach for diagnosis of Parkinson's
Disease (PD) based on deep Echo State Networks (ESNs). The identification of PD
is performed by analyzing the whole time-series collected from a tablet device
during the sketching of spiral tests, without the need for feature extraction
and data preprocessing. We evaluated the proposed approach on a public dataset
of spiral tests. The results of experimental analysis show that DeepESNs
perform significantly better than shallow ESN model. Overall, the proposed
approach obtains state-of-the-art results in the identification of PD on this
kind of temporal data.
|
Differentially Private Generative Adversarial Network | Generative Adversarial Network (GAN) and its variants have recently attracted
intensive research interests due to their elegant theoretical foundation and
excellent empirical performance as generative models. These tools provide a
promising direction in the studies where data availability is limited. One
common issue in GANs is that the density of the learned generative distribution
could concentrate on the training data points, meaning that they can easily
remember training samples due to the high model complexity of deep networks.
This becomes a major concern when GANs are applied to private or sensitive data
such as patient medical records, and the concentration of distribution may
divulge critical patient information. To address this issue, in this paper we
propose a differentially private GAN (DPGAN) model, in which we achieve
differential privacy in GANs by adding carefully designed noise to gradients
during the learning procedure. We provide rigorous proof for the privacy
guarantee, as well as comprehensive empirical evidence to support our analysis,
where we demonstrate that our method can generate high quality data points at a
reasonable privacy level.
|
Leveraged volume sampling for linear regression | Suppose an $n \times d$ design matrix in a linear regression problem is
given, but the response for each point is hidden unless explicitly requested.
The goal is to sample only a small number $k \ll n$ of the responses, and then
produce a weight vector whose sum of squares loss over all points is at most
$1+\epsilon$ times the minimum. When $k$ is very small (e.g., $k=d$), jointly
sampling diverse subsets of points is crucial. One such method called volume
sampling has a unique and desirable property that the weight vector it produces
is an unbiased estimate of the optimum. It is therefore natural to ask if this
method offers the optimal unbiased estimate in terms of the number of responses
$k$ needed to achieve a $1+\epsilon$ loss approximation.
Surprisingly we show that volume sampling can have poor behavior when we
require a very accurate approximation -- indeed worse than some i.i.d. sampling
techniques whose estimates are biased, such as leverage score sampling. We then
develop a new rescaled variant of volume sampling that produces an unbiased
estimate which avoids this bad behavior and has at least as good a tail bound
as leverage score sampling: sample size $k=O(d\log d + d/\epsilon)$ suffices to
guarantee total loss at most $1+\epsilon$ times the minimum with high
probability. Thus, we improve on the best previously known sample size for an
unbiased estimator, $k=O(d^2/\epsilon)$.
Our rescaling procedure leads to a new efficient algorithm for volume
sampling which is based on a determinantal rejection sampling technique with
potentially broader applications to determinantal point processes. Other
contributions include introducing the combinatorics needed for rescaled volume
sampling and developing tail bounds for sums of dependent random matrices which
arise in the process.
|
Interpretable VAEs for nonlinear group factor analysis | Deep generative models have recently yielded encouraging results in producing
subjectively realistic samples of complex data. Far less attention has been
paid to making these generative models interpretable. In many scenarios,
ranging from scientific applications to finance, the observed variables have a
natural grouping. It is often of interest to understand systems of interaction
amongst these groups, and latent factor models (LFMs) are an attractive
approach. However, traditional LFMs are limited by assuming a linear
correlation structure. We present an output interpretable VAE (oi-VAE) for
grouped data that models complex, nonlinear latent-to-observed relationships.
We combine a structured VAE comprised of group-specific generators with a
sparsity-inducing prior. We demonstrate that oi-VAE yields meaningful notions
of interpretability in the analysis of motion capture and MEG data. We further
show that in these situations, the regularization inherent to oi-VAE can
actually lead to improved generalization and learned generative processes.
|
Bayes-optimal Hierarchical Classification over Asymmetric Tree-Distance
Loss | Hierarchical classification is supervised multi-class classification problem
over the set of class labels organized according to a hierarchy. In this
report, we study the work by Ramaswamy et. al. on hierarchical classification
over symmetric tree distance loss. We extend the consistency of hierarchical
classification algorithm over asymmetric tree distance loss. We design a
$\mathcal{O}(nk\log{}n)$ algorithm to find Bayes optimal classification for a
k-ary tree as a hierarchy. We show that under reasonable assumptions over
asymmetric loss function, the Bayes optimal classification over this asymmetric
loss can be found in $\mathcal{O}(k\log{}n)$. We exploit this insight and
attempt to extend the Ova-Cascade algorithm \citet{ramaswamy2015convex} for
hierarchical classification over the asymmetric loss.
|
Entropy-Isomap: Manifold Learning for High-dimensional Dynamic Processes | Scientific and engineering processes deliver massive high-dimensional data
sets that are generated as non-linear transformations of an initial state and
few process parameters. Mapping such data to a low-dimensional manifold
facilitates better understanding of the underlying processes, and enables their
optimization. In this paper, we first show that off-the-shelf non-linear
spectral dimensionality reduction methods, e.g., Isomap, fail for such data,
primarily due to the presence of strong temporal correlations. Then, we propose
a novel method, Entropy-Isomap, to address the issue. The proposed method is
successfully applied to large data describing a fabrication process of organic
materials. The resulting low-dimensional representation correctly captures
process control variables, allows for low-dimensional visualization of the
material morphology evolution, and provides key insights to improve the
process.
|
Multi-resolution Tensor Learning for Large-Scale Spatial Data | High-dimensional tensor models are notoriously computationally expensive to
train. We present a meta-learning algorithm, MMT, that can significantly speed
up the process for spatial tensor models. MMT leverages the property that
spatial data can be viewed at multiple resolutions, which are related by
coarsening and finegraining from one resolution to another. Using this
property, MMT learns a tensor model by starting from a coarse resolution and
iteratively increasing the model complexity. In order to not "over-train" on
coarse resolution models, we investigate an information-theoretic fine-graining
criterion to decide when to transition into higher-resolution models. We
provide both theoretical and empirical evidence for the advantages of this
approach. When applied to two real-world large-scale spatial datasets for
basketball player and animal behavior modeling, our approach demonstrate 3 key
benefits: 1) it efficiently captures higher-order interactions (i.e., tensor
latent factors), 2) it is orders of magnitude faster than fixed resolution
learning and scales to very fine-grained spatial resolutions, and 3) it
reliably yields accurate and interpretable models.
|
Deep Learning for Joint Source-Channel Coding of Text | We consider the problem of joint source and channel coding of structured data
such as natural language over a noisy channel. The typical approach to this
problem in both theory and practice involves performing source coding to first
compress the text and then channel coding to add robustness for the
transmission across the channel. This approach is optimal in terms of
minimizing end-to-end distortion with arbitrarily large block lengths of both
the source and channel codes when transmission is over discrete memoryless
channels. However, the optimality of this approach is no longer ensured for
documents of finite length and limitations on the length of the encoding. We
will show in this scenario that we can achieve lower word error rates by
developing a deep learning based encoder and decoder. While the approach of
separate source and channel coding would minimize bit error rates, our approach
preserves semantic information of sentences by first embedding sentences in a
semantic space where sentences closer in meaning are located closer together,
and then performing joint source and channel coding on these embeddings.
|
Distribution Matching in Variational Inference | With the increasingly widespread deployment of generative models, there is a
mounting need for a deeper understanding of their behaviors and limitations. In
this paper, we expose the limitations of Variational Autoencoders (VAEs), which
consistently fail to learn marginal distributions in both latent and visible
spaces. We show this to be a consequence of learning by matching conditional
distributions, and the limitations of explicit model and posterior
distributions. It is popular to consider Generative Adversarial Networks (GANs)
as a means of overcoming these limitations, leading to hybrids of VAEs and
GANs. We perform a large-scale evaluation of several VAE-GAN hybrids and
analyze the implications of class probability estimation for learning
distributions. While promising, we conclude that at present, VAE-GAN hybrids
have limited applicability: they are harder to scale, evaluate, and use for
inference compared to VAEs; and they do not improve over the generation quality
of GANs.
|
Global Pose Estimation with an Attention-based Recurrent Network | The ability for an agent to localize itself within an environment is crucial
for many real-world applications. For unknown environments, Simultaneous
Localization and Mapping (SLAM) enables incremental and concurrent building of
and localizing within a map. We present a new, differentiable architecture,
Neural Graph Optimizer, progressing towards a complete neural network solution
for SLAM by designing a system composed of a local pose estimation model, a
novel pose selection module, and a novel graph optimization process. The entire
architecture is trained in an end-to-end fashion, enabling the network to
automatically learn domain-specific features relevant to the visual odometry
and avoid the involved process of feature engineering. We demonstrate the
effectiveness of our system on a simulated 2D maze and the 3D ViZ-Doom
environment.
|
LSALSA: Accelerated Source Separation via Learned Sparse Coding | We propose an efficient algorithm for the generalized sparse coding (SC)
inference problem. The proposed framework applies to both the single dictionary
setting, where each data point is represented as a sparse combination of the
columns of one dictionary matrix, as well as the multiple dictionary setting as
given in morphological component analysis (MCA), where the goal is to separate
a signal into additive parts such that each part has distinct sparse
representation within a corresponding dictionary. Both the SC task and its
generalization via MCA have been cast as $\ell_1$-regularized least-squares
optimization problems. To accelerate traditional acquisition of sparse codes,
we propose a deep learning architecture that constitutes a trainable
time-unfolded version of the Split Augmented Lagrangian Shrinkage Algorithm
(SALSA), a special case of the Alternating Direction Method of Multipliers
(ADMM). We empirically validate both variants of the algorithm, that we refer
to as LSALSA (learned-SALSA), on image vision tasks and demonstrate that at
inference our networks achieve vast improvements in terms of the running time,
the quality of estimated sparse codes, and visual clarity on both classic SC
and MCA problems. Finally, we present a theoretical framework for analyzing
LSALSA network: we show that the proposed approach exactly implements a
truncated ADMM applied to a new, learned cost function with curvature modified
by one of the learned parameterized matrices. We extend a very recent
Stochastic Alternating Optimization analysis framework to show that a gradient
descent step along this learned loss landscape is equivalent to a modified
gradient descent step along the original loss landscape. In this framework, the
acceleration achieved by LSALSA could potentially be explained by the network's
ability to learn a correction to the gradient direction of steeper descent.
|
Fourier Policy Gradients | We propose a new way of deriving policy gradient updates for reinforcement
learning. Our technique, based on Fourier analysis, recasts integrals that
arise with expected policy gradients as convolutions and turns them into
multiplications. The obtained analytical solutions allow us to capture the low
variance benefits of EPG in a broad range of settings. For the critic, we treat
trigonometric and radial basis functions, two function families with the
universal approximation property. The choice of policy can be almost arbitrary,
including mixtures or hybrid continuous-discrete probability distributions.
Moreover, we derive a general family of sample-based estimators for stochastic
policy gradients, which unifies existing results on sample-based approximation.
We believe that this technique has the potential to shape the next generation
of policy gradient approaches, powered by analytical results.
|
Learning Word Vectors for 157 Languages | Distributed word representations, or word vectors, have recently been applied
to many tasks in natural language processing, leading to state-of-the-art
performance. A key ingredient to the successful application of these
representations is to train them on very large corpora, and use these
pre-trained models in downstream tasks. In this paper, we describe how we
trained such high quality word representations for 157 languages. We used two
sources of data to train these models: the free online encyclopedia Wikipedia
and data from the common crawl project. We also introduce three new word
analogy datasets to evaluate these word vectors, for French, Hindi and Polish.
Finally, we evaluate our pre-trained word vectors on 10 languages for which
evaluation datasets exists, showing very strong performance compared to
previous models.
|
Learning Hidden Markov Models from Pairwise Co-occurrences with
Application to Topic Modeling | We present a new algorithm for identifying the transition and emission
probabilities of a hidden Markov model (HMM) from the emitted data.
Expectation-maximization becomes computationally prohibitive for long
observation records, which are often required for identification. The new
algorithm is particularly suitable for cases where the available sample size is
large enough to accurately estimate second-order output probabilities, but not
higher-order ones. We show that if one is only able to obtain a reliable
estimate of the pairwise co-occurrence probabilities of the emissions, it is
still possible to uniquely identify the HMM if the emission probability is
\emph{sufficiently scattered}. We apply our method to hidden topic Markov
modeling, and demonstrate that we can learn topics with higher quality if
documents are modeled as observations of HMMs sharing the same emission (topic)
probability, compared to the simple but widely used bag-of-words model.
|
Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative
Refinement | We propose a conditional non-autoregressive neural sequence model based on
iterative refinement. The proposed model is designed based on the principles of
latent variable models and denoising autoencoders, and is generally applicable
to any sequence generation task. We extensively evaluate the proposed model on
machine translation (En-De and En-Ro) and image caption generation, and observe
that it significantly speeds up decoding while maintaining the generation
quality comparable to the autoregressive counterpart.
|
Generalization Error Bounds with Probabilistic Guarantee for SGD in
Nonconvex Optimization | The success of deep learning has led to a rising interest in the
generalization property of the stochastic gradient descent (SGD) method, and
stability is one popular approach to study it. Existing works based on
stability have studied nonconvex loss functions, but only considered the
generalization error of the SGD in expectation. In this paper, we establish
various generalization error bounds with probabilistic guarantee for the SGD.
Specifically, for both general nonconvex loss functions and gradient dominant
loss functions, we characterize the on-average stability of the iterates
generated by SGD in terms of the on-average variance of the stochastic
gradients. Such characterization leads to improved bounds for the
generalization error for SGD. We then study the regularized risk minimization
problem with strongly convex regularizers, and obtain improved generalization
error bounds for proximal SGD. With strongly convex regularizers, we further
establish the generalization error bounds for nonconvex loss functions under
proximal SGD with high-probability guarantee, i.e., exponential concentration
in probability.
|
Teaching Categories to Human Learners with Visual Explanations | We study the problem of computer-assisted teaching with explanations.
Conventional approaches for machine teaching typically only provide feedback at
the instance level e.g., the category or label of the instance. However, it is
intuitive that clear explanations from a knowledgeable teacher can
significantly improve a student's ability to learn a new concept. To address
these existing limitations, we propose a teaching framework that provides
interpretable explanations as feedback and models how the learner incorporates
this additional information. In the case of images, we show that we can
automatically generate explanations that highlight the parts of the image that
are responsible for the class label. Experiments on human learners illustrate
that, on average, participants achieve better test set performance on
challenging categorization tasks when taught with our interpretable approach
compared to existing methods.
|
On Lyapunov exponents and adversarial perturbation | In this paper, we would like to disseminate a serendipitous discovery
involving Lyapunov exponents of a 1-D time series and their use in serving as a
filtering defense tool against a specific kind of deep adversarial
perturbation. To this end, we use the state-of-the-art CleverHans library to
generate adversarial perturbations against a standard Convolutional Neural
Network (CNN) architecture trained on the MNIST as well as the Fashion-MNIST
datasets. We empirically demonstrate how the Lyapunov exponents computed on the
flattened 1-D vector representations of the images served as highly
discriminative features that could be to pre-classify images as adversarial or
legitimate before feeding the image into the CNN for classification. We also
explore the issue of possible false-alarms when the input images are noisy in a
non-adversarial sense.
|
Online Learning with an Unknown Fairness Metric | We consider the problem of online learning in the linear contextual bandits
setting, but in which there are also strong individual fairness constraints
governed by an unknown similarity metric. These constraints demand that we
select similar actions or individuals with approximately equal probability
(arXiv:1104.3913), which may be at odds with optimizing reward, thus modeling
settings where profit and social policy are in tension. We assume we learn
about an unknown Mahalanobis similarity metric from only weak feedback that
identifies fairness violations, but does not quantify their extent. This is
intended to represent the interventions of a regulator who "knows unfairness
when he sees it" but nevertheless cannot enunciate a quantitative fairness
metric over individuals. Our main result is an algorithm in the adversarial
context setting that has a number of fairness violations that depends only
logarithmically on $T$, while obtaining an optimal $O(\sqrt{T})$ regret bound
to the best fair policy.
|
Estimator of Prediction Error Based on Approximate Message Passing for
Penalized Linear Regression | We propose an estimator of prediction error using an approximate message
passing (AMP) algorithm that can be applied to a broad range of sparse
penalties. Following Stein's lemma, the estimator of the generalized degrees of
freedom, which is a key quantity for the construction of the estimator of the
prediction error, is calculated at the AMP fixed point. The resulting form of
the AMP-based estimator does not depend on the penalty function, and its value
can be further improved by considering the correlation between predictors. The
proposed estimator is asymptotically unbiased when the components of the
predictors and response variables are independently generated according to a
Gaussian distribution. We examine the behaviour of the estimator for real data
under nonconvex sparse penalties, where Akaike's information criterion does not
correspond to an unbiased estimator of the prediction error. The model selected
by the proposed estimator is close to that which minimizes the true prediction
error.
|
Comparison Based Learning from Weak Oracles | There is increasing interest in learning algorithms that involve interaction
between human and machine. Comparison-based queries are among the most natural
ways to get feedback from humans. A challenge in designing comparison-based
interactive learning algorithms is coping with noisy answers. The most common
fix is to submit a query several times, but this is not applicable in many
situations due to its prohibitive cost and due to the unrealistic assumption of
independent noise in different repetitions of the same query.
In this paper, we introduce a new weak oracle model, where a non-malicious
user responds to a pairwise comparison query only when she is quite sure about
the answer. This model is able to mimic the behavior of a human in noise-prone
regions. We also consider the application of this weak oracle model to the
problem of content search (a variant of the nearest neighbor search problem)
through comparisons. More specifically, we aim at devising efficient algorithms
to locate a target object in a database equipped with a dissimilarity metric
via invocation of the weak comparison oracle. We propose two algorithms termed
WORCS-I and WORCS-II (Weak-Oracle Comparison-based Search), which provably
locate the target object in a number of comparisons close to the entropy of the
target distribution. While WORCS-I provides better theoretical guarantees,
WORCS-II is applicable to more technically challenging scenarios where the
algorithm has limited access to the ranking dissimilarity between objects. A
series of experiments validate the performance of our proposed algorithms.
|
DeepThin: A Self-Compressing Library for Deep Neural Networks | As the industry deploys increasingly large and complex neural networks to
mobile devices, more pressure is put on the memory and compute resources of
those devices. Deep compression, or compression of deep neural network weight
matrices, is a technique to stretch resources for such scenarios. Existing
compression methods cannot effectively compress models smaller than 1-2% of
their original size. We develop a new compression technique, DeepThin, building
on existing research in the area of low rank factorization. We identify and
break artificial constraints imposed by low rank approximations by combining
rank factorization with a reshaping process that adds nonlinearity to the
approximation function. We deploy DeepThin as a plug-gable library integrated
with TensorFlow that enables users to seamlessly compress models at different
granularities. We evaluate DeepThin on two state-of-the-art acoustic models,
TFKaldi and DeepSpeech, comparing it to previous compression work (Pruning,
HashNet, and Rank Factorization), empirical limit study approaches, and
hand-tuned models. For TFKaldi, our DeepThin networks show better word error
rates (WER) than competing methods at practically all tested compression rates,
achieving an average of 60% relative improvement over rank factorization, 57%
over pruning, 23% over hand-tuned same-size networks, and 6% over the
computationally expensive HashedNets. For DeepSpeech, DeepThin-compressed
networks achieve better test loss than all other compression methods, reaching
a 28% better result than rank factorization, 27% better than pruning, 20%
better than hand-tuned same-size networks, and 12% better than HashedNets.
DeepThin also provide inference performance benefits ranging from 2X to 14X
speedups, depending on the compression ratio and platform cache sizes.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.