title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Asymmetric Actor Critic for Image-Based Robot Learning | Deep reinforcement learning (RL) has proven a powerful technique in many
sequential decision making domains. However, Robotics poses many challenges for
RL, most notably training on a physical system can be expensive and dangerous,
which has sparked significant interest in learning control policies using a
physics simulator. While several recent works have shown promising results in
transferring policies trained in simulation to the real world, they often do
not fully utilize the advantage of working with a simulator. In this work, we
exploit the full state observability in the simulator to train better policies
which take as input only partial observations (RGBD images). We do this by
employing an actor-critic training algorithm in which the critic is trained on
full states while the actor (or policy) gets rendered images as input. We show
experimentally on a range of simulated tasks that using these asymmetric inputs
significantly improves performance. Finally, we combine this method with domain
randomization and show real robot experiments for several tasks like picking,
pushing, and moving a block. We achieve this simulation to real world transfer
without training on any real world data.
|
Replacement AutoEncoder: A Privacy-Preserving Algorithm for Sensory Data
Analysis | An increasing number of sensors on mobile, Internet of things (IoT), and
wearable devices generate time-series measurements of physical activities.
Though access to the sensory data is critical to the success of many beneficial
applications such as health monitoring or activity recognition, a wide range of
potentially sensitive information about the individuals can also be discovered
through access to sensory data and this cannot easily be protected using
traditional privacy approaches.
In this paper, we propose a privacy-preserving sensing framework for managing
access to time-series data in order to provide utility while protecting
individuals' privacy. We introduce Replacement AutoEncoder, a novel algorithm
which learns how to transform discriminative features of data that correspond
to sensitive inferences, into some features that have been more observed in
non-sensitive inferences, to protect users' privacy. This efficiency is
achieved by defining a user-customized objective function for deep
autoencoders. Our replacement method will not only eliminate the possibility of
recognizing sensitive inferences, it also eliminates the possibility of
detecting the occurrence of them. That is the main weakness of other approaches
such as filtering or randomization. We evaluate the efficacy of the algorithm
with an activity recognition task in a multi-sensing environment using
extensive experiments on three benchmark datasets. We show that it can retain
the recognition accuracy of state-of-the-art techniques while simultaneously
preserving the privacy of sensitive information. Finally, we utilize the GANs
for detecting the occurrence of replacement, after releasing data, and show
that this can be done only if the adversarial network is trained on the users'
original data.
|
A Correspondence Between Random Neural Networks and Statistical Field
Theory | A number of recent papers have provided evidence that practical design
questions about neural networks may be tackled theoretically by studying the
behavior of random networks. However, until now the tools available for
analyzing random neural networks have been relatively ad-hoc. In this work, we
show that the distribution of pre-activations in random neural networks can be
exactly mapped onto lattice models in statistical physics. We argue that
several previous investigations of stochastic networks actually studied a
particular factorial approximation to the full lattice model. For random linear
networks and random rectified linear networks we show that the corresponding
lattice models in the wide network limit may be systematically approximated by
a Gaussian distribution with covariance between the layers of the network. In
each case, the approximate distribution can be diagonalized by Fourier
transformation. We show that this approximation accurately describes the
results of numerical simulations of wide random neural networks. Finally, we
demonstrate that in each case the large scale behavior of the random networks
can be approximated by an effective field theory.
|
The Effects of Memory Replay in Reinforcement Learning | Experience replay is a key technique behind many recent advances in deep
reinforcement learning. Allowing the agent to learn from earlier memories can
speed up learning and break undesirable temporal correlations. Despite its
wide-spread application, very little is understood about the properties of
experience replay. How does the amount of memory kept affect learning dynamics?
Does it help to prioritize certain experiences? In this paper, we address these
questions by formulating a dynamical systems ODE model of Q-learning with
experience replay. We derive analytic solutions of the ODE for a simple
setting. We show that even in this very simple setting, the amount of memory
kept can substantially affect the agent's performance. Too much or too little
memory both slow down learning. Moreover, we characterize regimes where
prioritized replay harms the agent's learning. We show that our analytic
solutions have excellent agreement with experiments. Finally, we propose a
simple algorithm for adaptively changing the memory buffer size which achieves
consistently good empirical performance.
|
Learning Social Image Embedding with Deep Multimodal Attention Networks | Learning social media data embedding by deep models has attracted extensive
research interest as well as boomed a lot of applications, such as link
prediction, classification, and cross-modal search. However, for social images
which contain both link information and multimodal contents (e.g., text
description, and visual content), simply employing the embedding learnt from
network structure or data content results in sub-optimal social image
representation. In this paper, we propose a novel social image embedding
approach called Deep Multimodal Attention Networks (DMAN), which employs a deep
model to jointly embed multimodal contents and link information. Specifically,
to effectively capture the correlations between multimodal contents, we propose
a multimodal attention network to encode the fine-granularity relation between
image regions and textual words. To leverage the network structure for
embedding learning, a novel Siamese-Triplet neural network is proposed to model
the links among images. With the joint deep model, the learnt embedding can
capture both the multimodal contents and the nonlinear network information.
Extensive experiments are conducted to investigate the effectiveness of our
approach in the applications of multi-label classification and cross-modal
search. Compared to state-of-the-art image embeddings, our proposed DMAN
achieves significant improvement in the tasks of multi-label classification and
cross-modal search.
|
Function Norms and Regularization in Deep Networks | Deep neural networks (DNNs) have become increasingly important due to their
excellent empirical performance on a wide range of problems. However,
regularization is generally achieved by indirect means, largely due to the
complex set of functions defined by a network and the difficulty in measuring
function complexity. There exists no method in the literature for additive
regularization based on a norm of the function, as is classically considered in
statistical learning theory. In this work, we propose sampling-based
approximations to weighted function norms as regularizers for deep neural
networks. We provide, to the best of our knowledge, the first proof in the
literature of the NP-hardness of computing function norms of DNNs, motivating
the necessity of an approximate approach. We then derive a generalization bound
for functions trained with weighted norms and prove that a natural stochastic
optimization strategy minimizes the bound. Finally, we empirically validate the
improved performance of the proposed regularization strategies for both convex
function sets as well as DNNs on real-world classification and image
segmentation tasks demonstrating improved performance over weight decay,
dropout, and batch normalization. Source code will be released at the time of
publication.
|
A complete characterization of optimal dictionaries for least squares
representation | Dictionaries are collections of vectors used for representations of elements
in Euclidean spaces. While recent research on optimal dictionaries is focussed
on providing sparse (i.e., $\ell_0$-optimal,) representations, here we consider
the problem of finding optimal dictionaries such that representations of
samples of a random vector are optimal in an $\ell_2$-sense. For us, optimality
of representation is equivalent to minimization of the average $\ell_2$-norm of
the coefficients used to represent the random vector, with the lengths of the
dictionary vectors being specified a priori. With the help of recent results on
rank-$1$ decompositions of symmetric positive semidefinite matrices and the
theory of majorization, we provide a complete characterization of
$\ell_2$-optimal dictionaries. Our results are accompanied by polynomial time
algorithms that construct $\ell_2$-optimal dictionaries from given data.
|
Phase Transitions in the Pooled Data Problem | In this paper, we study the pooled data problem of identifying the labels
associated with a large collection of items, based on a sequence of pooled
tests revealing the counts of each label within the pool. In the noiseless
setting, we identify an exact asymptotic threshold on the required number of
tests with optimal decoding, and prove a phase transition between complete
success and complete failure. In addition, we present a novel noisy variation
of the problem, and provide an information-theoretic framework for
characterizing the required number of tests for general random noise models.
Our results reveal that noise can make the problem considerably more difficult,
with strict increases in the scaling laws even at low noise levels. Finally, we
demonstrate similar behavior in an approximate recovery setting, where a given
number of errors is allowed in the decoded labels.
|
Feature versus Raw Sequence: Deep Learning Comparative Study on
Predicting Pre-miRNA | Should we input known genome sequence features or input sequence itself in
deep learning framework? As deep learning more popular in various applications,
researchers often come to question whether to generate features or use raw
sequences for deep learning. To answer this question, we study the prediction
accuracy of precursor miRNA prediction of feature-based deep belief network and
sequence-based convolution neural network. Tested on a variant of six-layer
convolution neural net and three-layer deep belief network, we find the raw
sequence input based convolution neural network model performs similar or
slightly better than feature based deep belief networks with best accuracy
values of 0.995 and 0.990, respectively. Both the models outperform existing
benchmarks models. The results shows us that if provided large enough data,
well devised raw sequence based deep learning models can replace feature based
deep learning models. However, construction of well behaved deep learning model
can be very challenging. In cased features can be easily extracted,
feature-based deep learning models may be a better alternative.
|
Photo-Guided Exploration of Volume Data Features | In this work, we pose the question of whether, by considering qualitative
information such as a sample target image as input, one can produce a rendered
image of scientific data that is similar to the target. The algorithm resulting
from our research allows one to ask the question of whether features like those
in the target image exists in a given dataset. In that way, our method is one
of imagery query or reverse engineering, as opposed to manual parameter
tweaking of the full visualization pipeline. For target images, we can use
real-world photographs of physical phenomena. Our method leverages deep neural
networks and evolutionary optimization. Using a trained similarity function
that measures the difference between renderings of a phenomenon and real-world
photographs, our method optimizes rendering parameters. We demonstrate the
efficacy of our method using a superstorm simulation dataset and images found
online. We also discuss a parallel implementation of our method, which was run
on NCSA's Blue Waters.
|
The Origins of Computational Mechanics: A Brief Intellectual History and
Several Clarifications | The principle goal of computational mechanics is to define pattern and
structure so that the organization of complex systems can be detected and
quantified. Computational mechanics developed from efforts in the 1970s and
early 1980s to identify strange attractors as the mechanism driving weak fluid
turbulence via the method of reconstructing attractor geometry from measurement
time series and in the mid-1980s to estimate equations of motion directly from
complex time series. In providing a mathematical and operational definition of
structure it addressed weaknesses of these early approaches to discovering
patterns in natural systems.
Since then, computational mechanics has led to a range of results from
theoretical physics and nonlinear mathematics to diverse applications---from
closed-form analysis of Markov and non-Markov stochastic processes that are
ergodic or nonergodic and their measures of information and intrinsic
computation to complex materials and deterministic chaos and intelligence in
Maxwellian demons to quantum compression of classical processes and the
evolution of computation and language.
This brief review clarifies several misunderstandings and addresses concerns
recently raised regarding early works in the field (1980s). We show that
misguided evaluations of the contributions of computational mechanics are
groundless and stem from a lack of familiarity with its basic goals and from a
failure to consider its historical context. For all practical purposes, its
modern methods and results largely supersede the early works. This not only
renders recent criticism moot and shows the solid ground on which computational
mechanics stands but, most importantly, shows the significant progress achieved
over three decades and points to the many intriguing and outstanding challenges
in understanding the computational nature of complex dynamic systems.
|
Shannon Entropy Estimation in $\infty$-Alphabets from Convergence
Results | The problem of Shannon entropy estimation in countable infinite alphabets is
addressed from the study and use of convergence results of the entropy
functional, which is known to be discontinuous with respect to the total
variation distance in $\infty$-alphabets. Sufficient conditions for the
convergence of the entropy are used, including scenarios with both finitely and
infinitely supported assumptions on the distributions. From this new
perspective, four plug-in histogram-based estimators are studied showing that
convergence results are instrumental to derive new strong consistency and rate
of convergences results. Different scenarios and conditions are used on both
the estimators and the underlying distribution, considering for example finite
and unknown supported assumptions and summable tail bounded conditions.
|
Temporally-Reweighted Chinese Restaurant Process Mixtures for
Clustering, Imputing, and Forecasting Multivariate Time Series | This article proposes a Bayesian nonparametric method for forecasting,
imputation, and clustering in sparsely observed, multivariate time series data.
The method is appropriate for jointly modeling hundreds of time series with
widely varying, non-stationary dynamics. Given a collection of $N$ time series,
the Bayesian model first partitions them into independent clusters using a
Chinese restaurant process prior. Within a cluster, all time series are modeled
jointly using a novel "temporally-reweighted" extension of the Chinese
restaurant process mixture. Markov chain Monte Carlo techniques are used to
obtain samples from the posterior distribution, which are then used to form
predictive inferences. We apply the technique to challenging forecasting and
imputation tasks using seasonal flu data from the US Center for Disease Control
and Prevention, demonstrating superior forecasting accuracy and competitive
imputation accuracy as compared to multiple widely used baselines. We further
show that the model discovers interpretable clusters in datasets with hundreds
of time series, using macroeconomic data from the Gapminder Foundation.
|
Characterization of Gradient Dominance and Regularity Conditions for
Neural Networks | The past decade has witnessed a successful application of deep learning to
solving many challenging problems in machine learning and artificial
intelligence. However, the loss functions of deep neural networks (especially
nonlinear networks) are still far from being well understood from a theoretical
aspect. In this paper, we enrich the current understanding of the landscape of
the square loss functions for three types of neural networks. Specifically,
when the parameter matrices are square, we provide an explicit characterization
of the global minimizers for linear networks, linear residual networks, and
nonlinear networks with one hidden layer. Then, we establish two quadratic
types of landscape properties for the square loss of these neural networks,
i.e., the gradient dominance condition within the neighborhood of their full
rank global minimizers, and the regularity condition along certain directions
and within the neighborhood of their global minimizers. These two landscape
properties are desirable for the optimization around the global minimizers of
the loss function for these neural networks.
|
Concept Drift Learning with Alternating Learners | Data-driven predictive analytics are in use today across a number of
industrial applications, but further integration is hindered by the requirement
of similarity among model training and test data distributions. This paper
addresses the need of learning from possibly nonstationary data streams, or
under concept drift, a commonly seen phenomenon in practical applications. A
simple dual-learner ensemble strategy, alternating learners framework, is
proposed. A long-memory model learns stable concepts from a long relevant time
window, while a short-memory model learns transient concepts from a small
recent window. The difference in prediction performance of these two models is
monitored and induces an alternating policy to select, update and reset the two
models. The method features an online updating mechanism to maintain the
ensemble accuracy, and a concept-dependent trigger to focus on relevant data.
Through empirical studies the method demonstrates effective tracking and
prediction when the steaming data carry abrupt and/or gradual changes.
|
First-Person Perceptual Guidance Behavior Decomposition using Active
Constraint Classification | Humans exhibit a wide range of adaptive and robust dynamic motion behavior
that is yet unmatched by autonomous control systems. These capabilities are
essential for real-time behavior generation in cluttered environments. Recent
work suggests that human capabilities rely on task structure learning and
embedded or ecological cognition in the form of perceptual guidance. This paper
describes the experimental investigation of the functional elements of human
motion guidance, focusing on the control and perceptual mechanisms. The motion,
control, and perceptual data from first-person guidance experiments is
decomposed into elemental segments based on invariants. These elements are then
analyzed to determine their functional characteristics. The resulting model
explains the structure of the agent-environment interaction and provides lawful
descriptions of specific perceptual guidance and control mechanisms.
|
Asynchronous Decentralized Parallel Stochastic Gradient Descent | Most commonly used distributed machine learning systems are either
synchronous or centralized asynchronous. Synchronous algorithms like
AllReduce-SGD perform poorly in a heterogeneous environment, while asynchronous
algorithms using a parameter server suffer from 1) communication bottleneck at
parameter servers when workers are many, and 2) significantly worse convergence
when the traffic to parameter server is congested. Can we design an algorithm
that is robust in a heterogeneous environment, while being communication
efficient and maintaining the best-possible convergence rate? In this paper, we
propose an asynchronous decentralized stochastic gradient decent algorithm
(AD-PSGD) satisfying all above expectations. Our theoretical analysis shows
AD-PSGD converges at the optimal $O(1/\sqrt{K})$ rate as SGD and has linear
speedup w.r.t. number of workers. Empirically, AD-PSGD outperforms the best of
decentralized parallel SGD (D-PSGD), asynchronous parallel SGD (A-PSGD), and
standard data parallel SGD (AllReduce-SGD), often by orders of magnitude in a
heterogeneous environment. When training ResNet-50 on ImageNet with up to 128
GPUs, AD-PSGD converges (w.r.t epochs) similarly to the AllReduce-SGD, but each
epoch can be up to 4-8X faster than its synchronous counterparts in a
network-sharing HPC environment. To the best of our knowledge, AD-PSGD is the
first asynchronous algorithm that achieves a similar epoch-wise convergence
rate as AllReduce-SGD, at an over 100-GPU scale.
|
Learning Differentially Private Recurrent Language Models | We demonstrate that it is possible to train large recurrent language models
with user-level differential privacy guarantees with only a negligible cost in
predictive accuracy. Our work builds on recent advances in the training of deep
networks on user-partitioned data and privacy accounting for stochastic
gradient descent. In particular, we add user-level privacy protection to the
federated averaging algorithm, which makes "large step" updates from user-level
data. Our work demonstrates that given a dataset with a sufficiently large
number of users (a requirement easily met by even small internet-scale
datasets), achieving differential privacy comes at the cost of increased
computation, rather than in decreased utility as in most prior work. We find
that our private LSTM language models are quantitatively and qualitatively
similar to un-noised models when trained on a large dataset.
|
ProLanGO: Protein Function Prediction Using Neural~Machine Translation
Based on a Recurrent Neural Network | With the development of next generation sequencing techniques, it is fast and
cheap to determine protein sequences but relatively slow and expensive to
extract useful information from protein sequences because of limitations of
traditional biological experimental techniques. Protein function prediction has
been a long standing challenge to fill the gap between the huge amount of
protein sequences and the known function. In this paper, we propose a novel
method to convert the protein function problem into a language translation
problem by the new proposed protein sequence language "ProLan" to the protein
function language "GOLan", and build a neural machine translation model based
on recurrent neural networks to translate "ProLan" language to "GOLan"
language. We blindly tested our method by attending the latest third Critical
Assessment of Function Annotation (CAFA 3) in 2016, and also evaluate the
performance of our methods on selected proteins whose function was released
after CAFA competition. The good performance on the training and testing
datasets demonstrates that our new proposed method is a promising direction for
protein function prediction. In summary, we first time propose a method which
converts the protein function prediction problem to a language translation
problem and applies a neural machine translation model for protein function
prediction.
|
Meta-Learning via Feature-Label Memory Network | Deep learning typically requires training a very capable architecture using
large datasets. However, many important learning problems demand an ability to
draw valid inferences from small size datasets, and such problems pose a
particular challenge for deep learning. In this regard, various researches on
"meta-learning" are being actively conducted. Recent work has suggested a
Memory Augmented Neural Network (MANN) for meta-learning. MANN is an
implementation of a Neural Turing Machine (NTM) with the ability to rapidly
assimilate new data in its memory, and use this data to make accurate
predictions. In models such as MANN, the input data samples and their
appropriate labels from previous step are bound together in the same memory
locations. This often leads to memory interference when performing a task as
these models have to retrieve a feature of an input from a certain memory
location and read only the label information bound to that location. In this
paper, we tried to address this issue by presenting a more robust MANN. We
revisited the idea of meta-learning and proposed a new memory augmented neural
network by explicitly splitting the external memory into feature and label
memories. The feature memory is used to store the features of input data
samples and the label memory stores their labels. Hence, when predicting the
label of a given input, our model uses its feature memory unit as a reference
to extract the stored feature of the input, and based on that feature, it
retrieves the label information of the input from the label memory unit. In
order for the network to function in this framework, a new memory-writingmodule
to encode label information into the label memory in accordance with the
meta-learning task structure is designed. Here, we demonstrate that our model
outperforms MANN by a large margin in supervised one-shot classification tasks
using Omniglot and MNIST datasets.
|
Binary Classification from Positive-Confidence Data | Can we learn a binary classifier from only positive data, without any
negative data or unlabeled data? We show that if one can equip positive data
with confidence (positive-confidence), one can successfully learn a binary
classifier, which we name positive-confidence (Pconf) classification. Our work
is related to one-class classification which is aimed at "describing" the
positive class by clustering-related methods, but one-class classification does
not have the ability to tune hyper-parameters and their aim is not on
"discriminating" positive and negative classes. For the Pconf classification
problem, we provide a simple empirical risk minimization framework that is
model-independent and optimization-independent. We theoretically establish the
consistency and an estimation error bound, and demonstrate the usefulness of
the proposed method for training deep neural networks through experiments.
|
Informational Neurobayesian Approach to Neural Networks Training.
Opportunities and Prospects | A study of the classification problem in context of information theory is
presented in the paper. Current research in that field is focused on
optimisation and bayesian approach. Although that gives satisfying results,
they require a vast amount of data and computations to train on. Authors
propose a new concept named Informational Neurobayesian Approach (INA), which
allows to solve the same problems, but requires significantly less training
data as well as computational power. Experiments were conducted to compare its
performance with the traditional one and the results showed that capacity of
the INA is quite promising.
|
Machine Learning as Statistical Data Assimilation | We identify a strong equivalence between neural network based machine
learning (ML) methods and the formulation of statistical data assimilation
(DA), known to be a problem in statistical physics. DA, as used widely in
physical and biological sciences, systematically transfers information in
observations to a model of the processes producing the observations. The
correspondence is that layer label in the ML setting is the analog of time in
the data assimilation setting. Utilizing aspects of this equivalence we discuss
how to establish the global minimum of the cost functions in the ML context,
using a variational annealing method from DA. This provides a design method for
optimal networks for ML applications and may serve as the basis for
understanding the success of "deep learning". Results from an ML example are
presented.
When the layer label is taken to be continuous, the Euler-Lagrange equation
for the ML optimization problem is an ordinary differential equation, and we
see that the problem being solved is a two point boundary value problem. The
use of continuous layers is denoted "deepest learning". The Hamiltonian version
provides a direct rationale for back propagation as a solution method for the
canonical momentum; however, it suggests other solution methods are to be
preferred.
|
Decomposition of Uncertainty in Bayesian Deep Learning for Efficient and
Risk-sensitive Learning | Bayesian neural networks with latent variables are scalable and flexible
probabilistic models: They account for uncertainty in the estimation of the
network weights and, by making use of latent variables, can capture complex
noise patterns in the data. We show how to extract and decompose uncertainty
into epistemic and aleatoric components for decision-making purposes. This
allows us to successfully identify informative points for active learning of
functions with heteroscedastic and bimodal noise. Using the decomposition we
further define a novel risk-sensitive criterion for reinforcement learning to
identify policies that balance expected cost, model-bias and noise aversion.
|
Power Plant Performance Modeling with Concept Drift | Power plant is a complex and nonstationary system for which the traditional
machine learning modeling approaches fall short of expectations. The
ensemble-based online learning methods provide an effective way to continuously
learn from the dynamic environment and autonomously update models to respond to
environmental changes. This paper proposes such an online ensemble regression
approach to model power plant performance, which is critically important for
operation optimization. The experimental results on both simulated and real
data show that the proposed method can achieve performance with less than 1%
mean average percentage error, which meets the general expectations in field
operations.
|
Atypicality for Heart Rate Variability Using a Pattern-Tree Weighting
Method | Heart rate variability (HRV) is a vital measure of the autonomic nervous
system functionality and a key indicator of cardiovascular condition. This
paper proposes a novel method, called pattern tree which is an extension of
Willem's context tree to real-valued data, to investigate HRV via an
atypicality framework. In a previous paper atypicality was developed as method
for mining and discovery in "Big Data," which requires a universal approach.
Using the proposed pattern tree as a universal source coder in this framework
led to discovery of arrhythmias and unknown patterns in HRV Holter Monitoring.
|
Visual Integration of Data and Model Space in Ensemble Learning | Ensembles of classifier models typically deliver superior performance and can
outperform single classifier models given a dataset and classification task at
hand. However, the gain in performance comes together with the lack in
comprehensibility, posing a challenge to understand how each model affects the
classification outputs and where the errors come from. We propose a tight
visual integration of the data and the model space for exploring and combining
classifier models. We introduce a workflow that builds upon the visual
integration and enables the effective exploration of classification outputs and
models. We then present a use case in which we start with an ensemble
automatically selected by a standard ensemble selection algorithm, and show how
we can manipulate models and alternative combinations.
|
Scalable Gaussian Processes with Billions of Inducing Inputs via Tensor
Train Decomposition | We propose a method (TT-GP) for approximate inference in Gaussian Process
(GP) models. We build on previous scalable GP research including stochastic
variational inference based on inducing inputs, kernel interpolation, and
structure exploiting algebra. The key idea of our method is to use Tensor Train
decomposition for variational parameters, which allows us to train GPs with
billions of inducing inputs and achieve state-of-the-art results on several
benchmarks. Further, our approach allows for training kernels based on deep
neural networks without any modifications to the underlying GP model. A neural
network learns a multidimensional embedding for the data, which is used by the
GP to make the final prediction. We train GP and neural network parameters
end-to-end without pretraining, through maximization of GP marginal likelihood.
We show the efficiency of the proposed approach on several regression and
classification benchmark datasets including MNIST, CIFAR-10, and Airline.
|
Online Monotone Games | Algorithmic game theory (AGT) focuses on the design and analysis of
algorithms for interacting agents, with interactions rigorously formalized
within the framework of games. Results from AGT find applications in domains
such as online bidding auctions for web advertisements and network routing
protocols. Monotone games are games where agent strategies naturally converge
to an equilibrium state. Previous results in AGT have been obtained for convex,
socially-convex, or smooth games, but not monotone games. Our primary
theoretical contributions are defining the monotone game setting and its
extension to the online setting, a new notion of regret for this setting, and
accompanying algorithms that achieve sub-linear regret. We demonstrate the
utility of online monotone game theory on a variety of problem domains
including variational inequalities, reinforcement learning, and generative
adversarial networks.
|
Ligand Pose Optimization with Atomic Grid-Based Convolutional Neural
Networks | Docking is an important tool in computational drug discovery that aims to
predict the binding pose of a ligand to a target protein through a combination
of pose scoring and optimization. A scoring function that is differentiable
with respect to atom positions can be used for both scoring and gradient-based
optimization of poses for docking. Using a differentiable grid-based atomic
representation as input, we demonstrate that a scoring function learned by
training a convolutional neural network (CNN) to identify binding poses can
also be applied to pose optimization. We also show that an iteratively-trained
CNN that includes poses optimized by the first CNN in its training set performs
even better at optimizing randomly initialized poses than either the first CNN
scoring function or AutoDock Vina.
|
First-order Methods Almost Always Avoid Saddle Points | We establish that first-order methods avoid saddle points for almost all
initializations. Our results apply to a wide variety of first-order methods,
including gradient descent, block coordinate descent, mirror descent and
variants thereof. The connecting thread is that such algorithms can be studied
from a dynamical systems perspective in which appropriate instantiations of the
Stable Manifold Theorem allow for a global stability analysis. Thus, neither
access to second-order derivative information nor randomness beyond
initialization is necessary to provably avoid saddle points.
|
Differentially Private Empirical Risk Minimization with Input
Perturbation | We propose a novel framework for the differentially private ERM, input
perturbation. Existing differentially private ERM implicitly assumed that the
data contributors submit their private data to a database expecting that the
database invokes a differentially private mechanism for publication of the
learned model. In input perturbation, each data contributor independently
randomizes her/his data by itself and submits the perturbed data to the
database. We show that the input perturbation framework theoretically
guarantees that the model learned with the randomized data eventually satisfies
differential privacy with the prescribed privacy parameters. At the same time,
input perturbation guarantees that local differential privacy is guaranteed to
the server. We also show that the excess risk bound of the model learned with
input perturbation is $O(1/n)$ under a certain condition, where $n$ is the
sample size. This is the same as the excess risk bound of the state-of-the-art.
|
Multipartite Pooling for Deep Convolutional Neural Networks | We propose a novel pooling strategy that learns how to adaptively rank deep
convolutional features for selecting more informative representations. To this
end, we exploit discriminative analysis to project the features onto a space
spanned by the number of classes in the dataset under study. This maps the
notion of labels in the feature space into instances in the projected space. We
employ these projected distances as a measure to rank the existing features
with respect to their specific discriminant power for each individual class. We
then apply multipartite ranking to score the separability of the instances and
aggregate one-versus-all scores to compute an overall distinction score for
each feature. For the pooling, we pick features with the highest scores in a
pooling window instead of maximum, average or stochastic random assignments.
Our experiments on various benchmarks confirm that the proposed strategy of
multipartite pooling is highly beneficial to consistently improve the
performance of deep convolutional networks via better generalization of the
trained models for the test-time data.
|
Distributed Deep Transfer Learning by Basic Probability Assignment | Transfer learning is a popular practice in deep neural networks, but
fine-tuning of large number of parameters is a hard task due to the complex
wiring of neurons between splitting layers and imbalance distributions of data
in pretrained and transferred domains. The reconstruction of the original
wiring for the target domain is a heavy burden due to the size of
interconnections across neurons. We propose a distributed scheme that tunes the
convolutional filters individually while backpropagates them jointly by means
of basic probability assignment. Some of the most recent advances in evidence
theory show that in a vast variety of the imbalanced regimes, optimizing of
some proper objective functions derived from contingency matrices prevents
biases towards high-prior class distributions. Therefore, the original filters
get gradually transferred based on individual contributions to overall
performance of the target domain. This largely reduces the expected complexity
of transfer learning whilst highly improves precision. Our experiments on
standard benchmarks and scenarios confirm the consistent improvement of our
distributed deep transfer learning strategy.
|
Unified Backpropagation for Multi-Objective Deep Learning | A common practice in most of deep convolutional neural architectures is to
employ fully-connected layers followed by Softmax activation to minimize
cross-entropy loss for the sake of classification. Recent studies show that
substitution or addition of the Softmax objective to the cost functions of
support vector machines or linear discriminant analysis is highly beneficial to
improve the classification performance in hybrid neural networks. We propose a
novel paradigm to link the optimization of several hybrid objectives through
unified backpropagation. This highly alleviates the burden of extensive
boosting for independent objective functions or complex formulation of
multiobjective gradients. Hybrid loss functions are linked by basic probability
assignment from evidence theory. We conduct our experiments for a variety of
scenarios and standard datasets to evaluate the advantage of our proposed
unification approach to deliver consistent improvements into the classification
performance of deep convolutional neural networks.
|
Finite-dimensional Gaussian approximation with linear inequality
constraints | Introducing inequality constraints in Gaussian process (GP) models can lead
to more realistic uncertainties in learning a great variety of real-world
problems. We consider the finite-dimensional Gaussian approach from Maatouk and
Bay (2017) which can satisfy inequality conditions everywhere (either
boundedness, monotonicity or convexity). Our contributions are threefold.
First, we extend their approach in order to deal with general sets of linear
inequalities. Second, we explore several Markov Chain Monte Carlo (MCMC)
techniques to approximate the posterior distribution. Third, we investigate
theoretical and numerical properties of the constrained likelihood for
covariance parameter estimation. According to experiments on both artificial
and real data, our full framework together with a Hamiltonian Monte Carlo-based
sampler provides efficient results on both data fitting and uncertainty
quantification.
|
Learning Wasserstein Embeddings | The Wasserstein distance received a lot of attention recently in the
community of machine learning, especially for its principled way of comparing
distributions. It has found numerous applications in several hard problems,
such as domain adaptation, dimensionality reduction or generative models.
However, its use is still limited by a heavy computational cost. Our goal is to
alleviate this problem by providing an approximation mechanism that allows to
break its inherent complexity. It relies on the search of an embedding where
the Euclidean distance mimics the Wasserstein distance. We show that such an
embedding can be found with a siamese architecture associated with a decoder
network that allows to move from the embedding space back to the original input
space. Once this embedding has been found, computing optimization problems in
the Wasserstein space (e.g. barycenters, principal directions or even
archetypes) can be conducted extremely fast. Numerical experiments supporting
this idea are conducted on image datasets, and show the wide potential benefits
of our method.
|
Tracking the gradients using the Hessian: A new look at variance
reducing stochastic methods | Our goal is to improve variance reducing stochastic methods through better
control variates. We first propose a modification of SVRG which uses the
Hessian to track gradients over time, rather than to recondition, increasing
the correlation of the control variates and leading to faster theoretical
convergence close to the optimum. We then propose accurate and computationally
efficient approximations to the Hessian, both using a diagonal and a low-rank
matrix. Finally, we demonstrate the effectiveness of our method on a wide range
of problems.
|
HDR image reconstruction from a single exposure using deep CNNs | Camera sensors can only capture a limited range of luminance simultaneously,
and in order to create high dynamic range (HDR) images a set of different
exposures are typically combined. In this paper we address the problem of
predicting information that have been lost in saturated image areas, in order
to enable HDR reconstruction from a single exposure. We show that this problem
is well-suited for deep learning algorithms, and propose a deep convolutional
neural network (CNN) that is specifically designed taking into account the
challenges in predicting HDR values. To train the CNN we gather a large dataset
of HDR images, which we augment by simulating sensor saturation for a range of
cameras. To further boost robustness, we pre-train the CNN on a simulated HDR
dataset created from a subset of the MIT Places database. We demonstrate that
our approach can reconstruct high-resolution visually convincing HDR results in
a wide range of situations, and that it generalizes well to reconstruction of
images captured with arbitrary and low-end cameras that use unknown camera
response functions and post-processing. Furthermore, we compare to existing
methods for HDR expansion, and show high quality results also for image based
lighting. Finally, we evaluate the results in a subjective experiment performed
on an HDR display. This shows that the reconstructed HDR images are visually
convincing, with large improvements as compared to existing methods.
|
Dynamic classifier chains for multi-label learning | In this paper, we deal with the task of building a dynamic ensemble of chain
classifiers for multi-label classification. To do so, we proposed two concepts
of classifier chains algorithms that are able to change label order of the
chain without rebuilding the entire model. Such modes allows anticipating the
instance-specific chain order without a significant increase in computational
burden. The proposed chain models are built using the Naive Bayes classifier
and nearest neighbour approach as a base single-label classifiers. To take the
benefits of the proposed algorithms, we developed a simple heuristic that
allows the system to find relatively good label order. The heuristic sort
labels according to the label-specific classification quality gained during the
validation phase. The heuristic tries to minimise the phenomenon of error
propagation in the chain. The experimental results showed that the proposed
model based on Naive Bayes classifier the above-mentioned heuristic is an
efficient tool for building dynamic chain classifiers.
|
Data-Free Knowledge Distillation for Deep Neural Networks | Recent advances in model compression have provided procedures for compressing
large neural networks to a fraction of their original size while retaining most
if not all of their accuracy. However, all of these approaches rely on access
to the original training set, which might not always be possible if the network
to be compressed was trained on a very large dataset, or on a dataset whose
release poses privacy or safety concerns as may be the case for biometrics
tasks. We present a method for data-free knowledge distillation, which is able
to compress deep neural networks trained on large-scale datasets to a fraction
of their size leveraging only some extra metadata to be provided with a
pretrained model release. We also explore different kinds of metadata that can
be used with our method, and discuss tradeoffs involved in using each of them.
|
Learning compressed representations of blood samples time series with
missing data | Clinical measurements collected over time are naturally represented as
multivariate time series (MTS), which often contain missing data. An
autoencoder can learn low dimensional vectorial representations of MTS that
preserve important data characteristics, but cannot deal explicitly with
missing data. In this work, we propose a new framework that combines an
autoencoder with the Time series Cluster Kernel (TCK), a kernel that accounts
for missingness patterns in MTS. Via kernel alignment, we incorporate TCK in
the autoencoder to improve the learned representations in presence of missing
data. We consider a classification problem of MTS with missing values,
representing blood samples of patients with surgical site infection. With our
approach, rather than with a standard autoencoder, we learn representations in
low dimensions that can be classified better.
|
Deep Voice 3: Scaling Text-to-Speech with Convolutional Sequence
Learning | We present Deep Voice 3, a fully-convolutional attention-based neural
text-to-speech (TTS) system. Deep Voice 3 matches state-of-the-art neural
speech synthesis systems in naturalness while training ten times faster. We
scale Deep Voice 3 to data set sizes unprecedented for TTS, training on more
than eight hundred hours of audio from over two thousand speakers. In addition,
we identify common error modes of attention-based speech synthesis networks,
demonstrate how to mitigate them, and compare several different waveform
synthesis methods. We also describe how to scale inference to ten million
queries per day on one single-GPU server.
|
On the Consistency of Graph-based Bayesian Learning and the Scalability
of Sampling Algorithms | A popular approach to semi-supervised learning proceeds by endowing the input
data with a graph structure in order to extract geometric information and
incorporate it into a Bayesian framework. We introduce new theory that gives
appropriate scalings of graph parameters that provably lead to a well-defined
limiting posterior as the size of the unlabeled data set grows. Furthermore, we
show that these consistency results have profound algorithmic implications.
When consistency holds, carefully designed graph-based Markov chain Monte Carlo
algorithms are proved to have a uniform spectral gap, independent of the number
of unlabeled inputs. Several numerical experiments corroborate both the
statistical consistency and the algorithmic scalability established by the
theory.
|
Low Precision RNNs: Quantizing RNNs Without Losing Accuracy | Similar to convolution neural networks, recurrent neural networks (RNNs)
typically suffer from over-parameterization. Quantizing bit-widths of weights
and activations results in runtime efficiency on hardware, yet it often comes
at the cost of reduced accuracy. This paper proposes a quantization approach
that increases model size with bit-width reduction. This approach will allow
networks to perform at their baseline accuracy while still maintaining the
benefits of reduced precision and overall model size reduction.
|
A Tight Excess Risk Bound via a Unified
PAC-Bayesian-Rademacher-Shtarkov-MDL Complexity | We present a novel notion of complexity that interpolates between and
generalizes some classic existing complexity notions in learning theory: for
estimators like empirical risk minimization (ERM) with arbitrary bounded
losses, it is upper bounded in terms of data-independent Rademacher complexity;
for generalized Bayesian estimators, it is upper bounded by the data-dependent
information complexity (also known as stochastic or PAC-Bayesian,
$\mathrm{KL}(\text{posterior} \operatorname{\|} \text{prior})$ complexity. For
(penalized) ERM, the new complexity reduces to (generalized) normalized maximum
likelihood (NML) complexity, i.e. a minimax log-loss individual-sequence
regret. Our first main result bounds excess risk in terms of the new
complexity. Our second main result links the new complexity via Rademacher
complexity to $L_2(P)$ entropy, thereby generalizing earlier results of Opper,
Haussler, Lugosi, and Cesa-Bianchi who did the log-loss case with $L_\infty$.
Together, these results recover optimal bounds for VC- and large (polynomial
entropy) classes, replacing localized Rademacher complexity by a simpler
analysis which almost completely separates the two aspects that determine the
achievable rates: 'easiness' (Bernstein) conditions and model complexity.
|
Learning Discrete Weights Using the Local Reparameterization Trick | Recent breakthroughs in computer vision make use of large deep neural
networks, utilizing the substantial speedup offered by GPUs. For applications
running on limited hardware, however, high precision real-time processing can
still be a challenge. One approach to solving this problem is training networks
with binary or ternary weights, thus removing the need to calculate
multiplications and significantly reducing memory size. In this work, we
introduce LR-nets (Local reparameterization networks), a new method for
training neural networks with discrete weights using stochastic parameters. We
show how a simple modification to the local reparameterization trick,
previously used to train Gaussian distributed weights, enables the training of
discrete weights. Using the proposed training we test both binary and ternary
models on MNIST, CIFAR-10 and ImageNet benchmarks and reach state-of-the-art
results on most experiments.
|
Towards Black-box Iterative Machine Teaching | In this paper, we make an important step towards the black-box machine
teaching by considering the cross-space machine teaching, where the teacher and
the learner use different feature representations and the teacher can not fully
observe the learner's model. In such scenario, we study how the teacher is
still able to teach the learner to achieve faster convergence rate than the
traditional passive learning. We propose an active teacher model that can
actively query the learner (i.e., make the learner take exams) for estimating
the learner's status and provably guide the learner to achieve faster
convergence. The sample complexities for both teaching and query are provided.
In the experiments, we compare the proposed active teacher with the omniscient
teacher and verify the effectiveness of the active teacher model.
|
Stochastic Backward Euler: An Implicit Gradient Descent Algorithm for
$k$-means Clustering | In this paper, we propose an implicit gradient descent algorithm for the
classic $k$-means problem. The implicit gradient step or backward Euler is
solved via stochastic fixed-point iteration, in which we randomly sample a
mini-batch gradient in every iteration. It is the average of the fixed-point
trajectory that is carried over to the next gradient step. We draw connections
between the proposed stochastic backward Euler and the recent entropy
stochastic gradient descent (Entropy-SGD) for improving the training of deep
neural networks. Numerical experiments on various synthetic and real datasets
show that the proposed algorithm provides better clustering results compared to
$k$-means algorithms in the sense that it decreased the objective function (the
cluster) and is much more robust to initialization.
|
A Novel Stochastic Stratified Average Gradient Method: Convergence Rate
and Its Complexity | SGD (Stochastic Gradient Descent) is a popular algorithm for large scale
optimization problems due to its low iterative cost. However, SGD can not
achieve linear convergence rate as FGD (Full Gradient Descent) because of the
inherent gradient variance. To attack the problem, mini-batch SGD was proposed
to get a trade-off in terms of convergence rate and iteration cost. In this
paper, a general CVI (Convergence-Variance Inequality) equation is presented to
state formally the interaction of convergence rate and gradient variance. Then
a novel algorithm named SSAG (Stochastic Stratified Average Gradient) is
introduced to reduce gradient variance based on two techniques, stratified
sampling and averaging over iterations that is a key idea in SAG (Stochastic
Average Gradient). Furthermore, SSAG can achieve linear convergence rate of
$\mathcal {O}((1-\frac{\mu}{8CL})^k)$ at smaller storage and iterative costs,
where $C\geq 2$ is the category number of training data. This convergence rate
depends mainly on the variance between classes, but not on the variance within
the classes. In the case of $C\ll N$ ($N$ is the training data size), SSAG's
convergence rate is much better than SAG's convergence rate of $\mathcal
{O}((1-\frac{\mu}{8NL})^k)$. Our experimental results show SSAG outperforms SAG
and many other algorithms.
|
Optimal Rates for Learning with Nystr\"om Stochastic Gradient Methods | In the setting of nonparametric regression, we propose and study a
combination of stochastic gradient methods with Nystr\"om subsampling, allowing
multiple passes over the data and mini-batches. Generalization error bounds for
the studied algorithm are provided. Particularly, optimal learning rates are
derived considering different possible choices of the step-size, the mini-batch
size, the number of iterations/passes, and the subsampling level. In comparison
with state-of-the-art algorithms such as the classic stochastic gradient
methods and kernel ridge regression with Nystr\"om, the studied algorithm has
advantages on the computational complexity, while achieving the same optimal
learning rates. Moreover, our results indicate that using mini-batches can
reduce the total computational cost while achieving the same optimal
statistical results.
|
Zeroth-Order Online Alternating Direction Method of Multipliers:
Convergence Analysis and Applications | In this paper, we design and analyze a new zeroth-order online algorithm,
namely, the zeroth-order online alternating direction method of multipliers
(ZOO-ADMM), which enjoys dual advantages of being gradient-free operation and
employing the ADMM to accommodate complex structured regularizers. Compared to
the first-order gradient-based online algorithm, we show that ZOO-ADMM requires
$\sqrt{m}$ times more iterations, leading to a convergence rate of
$O(\sqrt{m}/\sqrt{T})$, where $m$ is the number of optimization variables, and
$T$ is the number of iterations. To accelerate ZOO-ADMM, we propose two
minibatch strategies: gradient sample averaging and observation averaging,
resulting in an improved convergence rate of $O(\sqrt{1+q^{-1}m}/\sqrt{T})$,
where $q$ is the minibatch size. In addition to convergence analysis, we also
demonstrate ZOO-ADMM to applications in signal processing, statistics, and
machine learning.
|
A Learning-to-Infer Method for Real-Time Power Grid Multi-Line Outage
Identification | Identifying a potentially large number of simultaneous line outages in power
transmission networks in real time is a computationally hard problem. This is
because the number of hypotheses grows exponentially with the network size. A
new "Learning-to-Infer" method is developed for efficient inference of every
line status in the network. Optimizing the line outage detector is transformed
to and solved as a discriminative learning problem based on Monte Carlo samples
generated with power flow simulations. A major advantage of the developed
Learning-to-Infer method is that the labeled data used for training can be
generated in an arbitrarily large amount rapidly and at very little cost. As a
result, the power of offline training is fully exploited to learn very complex
classifiers for effective real-time multi-line outage identification. The
proposed methods are evaluated in the IEEE 30, 118 and 300 bus systems.
Excellent performance in identifying multi-line outages in real time is
achieved with a reasonably small amount of data.
|
Incomplete Dot Products for Dynamic Computation Scaling in Neural
Network Inference | We propose the use of incomplete dot products (IDP) to dynamically adjust the
number of input channels used in each layer of a convolutional neural network
during feedforward inference. IDP adds monotonically non-increasing
coefficients, referred to as a "profile", to the channels during training. The
profile orders the contribution of each channel in non-increasing order. At
inference time, the number of channels used can be dynamically adjusted to
trade off accuracy for lowered power consumption and reduced latency by
selecting only a beginning subset of channels. This approach allows for a
single network to dynamically scale over a computation range, as opposed to
training and deploying multiple networks to support different levels of
computation scaling. Additionally, we extend the notion to multiple profiles,
each optimized for some specific range of computation scaling. We present
experiments on the computation and accuracy trade-offs of IDP for popular image
classification models and datasets. We demonstrate that, for MNIST and
CIFAR-10, IDP reduces computation significantly, e.g., by 75%, without
significantly compromising accuracy. We argue that IDP provides a convenient
and effective means for devices to lower computation costs dynamically to
reflect the current computation budget of the system. For example, VGG-16 with
50% IDP (using only the first 50% of channels) achieves 70% in accuracy on the
CIFAR-10 dataset compared to the standard network which achieves only 35%
accuracy when using the reduced channel set.
|
A Generative Restricted Boltzmann Machine Based Method for
High-Dimensional Motion Data Modeling | Many computer vision applications involve modeling complex spatio-temporal
patterns in high-dimensional motion data. Recently, restricted Boltzmann
machines (RBMs) have been widely used to capture and represent spatial patterns
in a single image or temporal patterns in several time slices. To model global
dynamics and local spatial interactions, we propose to theoretically extend the
conventional RBMs by introducing another term in the energy function to
explicitly model the local spatial interactions in the input data. A learning
method is then proposed to perform efficient learning for the proposed model.
We further introduce a new method for multi-class classification that can
effectively estimate the infeasible partition functions of different RBMs such
that RBM is treated as a generative model for classification purpose. The
improved RBM model is evaluated on two computer vision applications: facial
expression recognition and human action recognition. Experimental results on
benchmark databases demonstrate the effectiveness of the proposed algorithm.
|
Deep Neural Network Approximation using Tensor Sketching | Deep neural networks are powerful learning models that achieve
state-of-the-art performance on many computer vision, speech, and language
processing tasks. In this paper, we study a fundamental question that arises
when designing deep network architectures: Given a target network architecture
can we design a smaller network architecture that approximates the operation of
the target network? The question is, in part, motivated by the challenge of
parameter reduction (compression) in modern deep neural networks, as the ever
increasing storage and memory requirements of these networks pose a problem in
resource constrained environments.
In this work, we focus on deep convolutional neural network architectures,
and propose a novel randomized tensor sketching technique that we utilize to
develop a unified framework for approximating the operation of both the
convolutional and fully connected layers. By applying the sketching technique
along different tensor dimensions, we design changes to the convolutional and
fully connected layers that substantially reduce the number of effective
parameters in a network. We show that the resulting smaller network can be
trained directly, and has a classification accuracy that is comparable to the
original network.
|
Deep Triphone Embedding Improves Phoneme Recognition | In this paper, we present a novel Deep Triphone Embedding (DTE)
representation derived from Deep Neural Network (DNN) to encapsulate the
discriminative information present in the adjoining speech frames. DTEs are
generated using a four hidden layer DNN with 3000 nodes in each hidden layer at
the first-stage. This DNN is trained with the tied-triphone classification
accuracy as an optimization criterion. Thereafter, we retain the activation
vectors (3000) of the last hidden layer, for each speech MFCC frame, and
perform dimension reduction to further obtain a 300 dimensional representation,
which we termed as DTE. DTEs along with MFCC features are fed into a
second-stage four hidden layer DNN, which is subsequently trained for the task
of tied-triphone classification. Both DNNs are trained using tri-phone labels
generated from a tied-state triphone HMM-GMM system, by performing a
forced-alignment between the transcriptions and MFCC feature frames. We conduct
the experiments on publicly available TED-LIUM speech corpus. The results show
that the proposed DTE method provides an improvement of absolute 2.11% in
phoneme recognition, when compared with a competitive hybrid tied-state
triphone HMM-DNN system.
|
Strategic Classification from Revealed Preferences | We study an online linear classification problem, in which the data is
generated by strategic agents who manipulate their features in an effort to
change the classification outcome. In rounds, the learner deploys a classifier,
and an adversarially chosen agent arrives, possibly manipulating her features
to optimally respond to the learner. The learner has no knowledge of the
agents' utility functions or "real" features, which may vary widely across
agents. Instead, the learner is only able to observe their "revealed
preferences" --- i.e. the actual manipulated feature vectors they provide. For
a broad family of agent cost functions, we give a computationally efficient
learning algorithm that is able to obtain diminishing "Stackelberg regret" ---
a form of policy regret that guarantees that the learner is obtaining loss
nearly as small as that of the best classifier in hindsight, even allowing for
the fact that agents will best-respond differently to the optimal classifier.
|
Bayesian Cluster Enumeration Criterion for Unsupervised Learning | We derive a new Bayesian Information Criterion (BIC) by formulating the
problem of estimating the number of clusters in an observed data set as
maximization of the posterior probability of the candidate models. Given that
some mild assumptions are satisfied, we provide a general BIC expression for a
broad class of data distributions. This serves as a starting point when
deriving the BIC for specific distributions. Along this line, we provide a
closed-form BIC expression for multivariate Gaussian distributed variables. We
show that incorporating the data structure of the clustering problem into the
derivation of the BIC results in an expression whose penalty term is different
from that of the original BIC. We propose a two-step cluster enumeration
algorithm. First, a model-based unsupervised learning algorithm partitions the
data according to a given set of candidate models. Subsequently, the number of
clusters is determined as the one associated with the model for which the
proposed BIC is maximal. The performance of the proposed two-step algorithm is
tested using synthetic and real data sets.
|
Rethinking Convolutional Semantic Segmentation Learning | Deep convolutional semantic segmentation (DCSS) learning doesn't converge to
an optimal local minimum with random parameters initializations; a pre-trained
model on the same domain becomes necessary to achieve convergence.In this work,
we propose a joint cooperative end-to-end learning method for DCSS. It
addresses many drawbacks with existing deep semantic segmentation learning; the
proposed approach simultaneously learn both segmentation and classification;
taking away the essential need of the pre-trained model for learning
convergence. We present an improved inception based architecture with partial
attention gating (PAG) over encoder information. The PAG also adds to achieve
faster convergence and better accuracy for segmentation task. We will show the
effectiveness of this learning on a diabetic retinopathy classification and
segmentation dataset.
|
Smart "Predict, then Optimize" | Many real-world analytics problems involve two significant challenges:
prediction and optimization. Due to the typically complex nature of each
challenge, the standard paradigm is predict-then-optimize. By and large,
machine learning tools are intended to minimize prediction error and do not
account for how the predictions will be used in the downstream optimization
problem. In contrast, we propose a new and very general framework, called Smart
"Predict, then Optimize" (SPO), which directly leverages the optimization
problem structure, i.e., its objective and constraints, for designing better
prediction models. A key component of our framework is the SPO loss function
which measures the decision error induced by a prediction.
Training a prediction model with respect to the SPO loss is computationally
challenging, and thus we derive, using duality theory, a convex surrogate loss
function which we call the SPO+ loss. Most importantly, we prove that the SPO+
loss is statistically consistent with respect to the SPO loss under mild
conditions. Our SPO+ loss function can tractably handle any polyhedral, convex,
or even mixed-integer optimization problem with a linear objective. Numerical
experiments on shortest path and portfolio optimization problems show that the
SPO framework can lead to significant improvement under the
predict-then-optimize paradigm, in particular when the prediction model being
trained is misspecified. We find that linear models trained using SPO+ loss
tend to dominate random forest algorithms, even when the ground truth is highly
nonlinear.
|
Exploiting generalization in the subspaces for faster model-based
learning | Due to the lack of enough generalization in the state-space, common methods
in Reinforcement Learning (RL) suffer from slow learning speed especially in
the early learning trials. This paper introduces a model-based method in
discrete state-spaces for increasing learning speed in terms of required
experience (but not required computational time) by exploiting generalization
in the experiences of the subspaces. A subspace is formed by choosing a subset
of features in the original state representation (full-space). Generalization
and faster learning in a subspace are due to many-to-one mapping of experiences
from the full-space to each state in the subspace. Nevertheless, due to
inherent perceptual aliasing in the subspaces, the policy suggested by each
subspace does not generally converge to the optimal policy. Our approach,
called Model Based Learning with Subspaces (MoBLeS), calculates confidence
intervals of the estimated Q-values in the full-space and in the subspaces.
These confidence intervals are used in the decision making, such that the agent
benefits the most from the possible generalization while avoiding from
detriment of the perceptual aliasing in the subspaces. Convergence of MoBLeS to
the optimal policy is theoretically investigated. Additionally, we show through
several experiments that MoBLeS improves the learning speed in the early
trials.
|
Sequential Matrix Completion | We propose a novel algorithm for sequential matrix completion in a
recommender system setting, where the $(i,j)$th entry of the matrix corresponds
to a user $i$'s rating of product $j$. The objective of the algorithm is to
provide a sequential policy for user-product pair recommendation which will
yield the highest possible ratings after a finite time horizon. The algorithm
uses a Gamma process factor model with two posterior-focused bandit policies,
Thompson Sampling and Information-Directed Sampling. While Thompson Sampling
shows competitive performance in simulations, state-of-the-art performance is
obtained from Information-Directed Sampling, which makes its recommendations
based off a ratio between the expected reward and a measure of information
gain. To our knowledge, this is the first implementation of Information
Directed Sampling on large real datasets.
This approach contributes to a recent line of research on bandit approaches
to collaborative filtering including Kawale et al. (2015), Li et al. (2010),
Bresler et al. (2014), Li et al. (2016), Deshpande & Montanari (2012), and Zhao
et al. (2013). The setting of this paper, as has been noted in Kawale et al.
(2015) and Zhao et al. (2013), presents significant challenges to bounding
regret after finite horizons. We discuss these challenges in relation to
simpler models for bandits with side information, such as linear or gaussian
process bandits, and hope the experiments presented here motivate further
research toward theoretical guarantees.
|
Accelerated Reinforcement Learning | Policy gradient methods are widely used in reinforcement learning algorithms
to search for better policies in the parameterized policy space. They do
gradient search in the policy space and are known to converge very slowly.
Nesterov developed an accelerated gradient search algorithm for convex
optimization problems. This has been recently extended for non-convex and also
stochastic optimization. We use Nesterov's acceleration for policy gradient
search in the well-known actor-critic algorithm and show the convergence using
ODE method. We tested this algorithm on a scheduling problem. Here an incoming
job is scheduled into one of the four queues based on the queue lengths. We see
from experimental results that algorithm using Nesterov's acceleration has
significantly better performance compared to algorithm which do not use
acceleration. To the best of our knowledge this is the first time Nesterov's
acceleration has been used with actor-critic algorithm.
|
Online Boosting Algorithms for Multi-label Ranking | We consider the multi-label ranking approach to multi-label learning.
Boosting is a natural method for multi-label ranking as it aggregates weak
predictions through majority votes, which can be directly used as scores to
produce a ranking of the labels. We design online boosting algorithms with
provable loss bounds for multi-label ranking. We show that our first algorithm
is optimal in terms of the number of learners required to attain a desired
accuracy, but it requires knowledge of the edge of the weak learners. We also
design an adaptive algorithm that does not require this knowledge and is hence
more practical. Experimental results on real data sets demonstrate that our
algorithms are at least as good as existing batch boosting algorithms.
|
Aggregating Algorithm for Prediction of Packs | This paper formulates the protocol for prediction of packs, which a special
case of prediction under delayed feedback. Under this protocol, the learner
must make a few predictions without seeing the outcomes and then the outcomes
are revealed. We develop the theory of prediction with expert advice for packs.
By applying Vovk's Aggregating Algorithm to this problem we obtain a number of
algorithms with tight upper bounds. We carry out empirical experiments on
housing data.
|
Interactive Visual Data Exploration with Subjective Feedback: An
Information-Theoretic Approach | Visual exploration of high-dimensional real-valued datasets is a fundamental
task in exploratory data analysis (EDA). Existing methods use predefined
criteria to choose the representation of data. There is a lack of methods that
(i) elicit from the user what she has learned from the data and (ii) show
patterns that she does not know yet. We construct a theoretical model where
identified patterns can be input as knowledge to the system. The knowledge
syntax here is intuitive, such as "this set of points forms a cluster", and
requires no knowledge of maths. This background knowledge is used to find a
Maximum Entropy distribution of the data, after which the system provides the
user data projections in which the data and the Maximum Entropy distribution
differ the most, hence showing the user aspects of the data that are maximally
informative given the user's current knowledge. We provide an open source EDA
system with tailored interactive visualizations to demonstrate these concepts.
We study the performance of the system and present use cases on both synthetic
and real data. We find that the model and the prototype system allow the user
to learn information efficiently from various data sources and the system works
sufficiently fast in practice. We conclude that the information theoretic
approach to exploratory data analysis where patterns observed by a user are
formalized as constraints provides a principled, intuitive, and efficient basis
for constructing an EDA system.
|
Progressive Learning for Systematic Design of Large Neural Networks | We develop an algorithm for systematic design of a large artificial neural
network using a progression property. We find that some non-linear functions,
such as the rectifier linear unit and its derivatives, hold the property. The
systematic design addresses the choice of network size and regularization of
parameters. The number of nodes and layers in network increases in progression
with the objective of consistently reducing an appropriate cost. Each layer is
optimized at a time, where appropriate parameters are learned using convex
optimization. Regularization parameters for convex optimization do not need a
significant manual effort for tuning. We also use random instances for some
weight matrices, and that helps to reduce the number of parameters we learn.
The developed network is expected to show good generalization power due to
appropriate regularization and use of random weights in the layers. This
expectation is verified by extensive experiments for classification and
regression problems, using standard databases.
|
Generic 3D Representation via Pose Estimation and Matching | Though a large body of computer vision research has investigated developing
generic semantic representations, efforts towards developing a similar
representation for 3D has been limited. In this paper, we learn a generic 3D
representation through solving a set of foundational proxy 3D tasks:
object-centric camera pose estimation and wide baseline feature matching. Our
method is based upon the premise that by providing supervision over a set of
carefully selected foundational tasks, generalization to novel tasks and
abstraction capabilities can be achieved. We empirically show that the internal
representation of a multi-task ConvNet trained to solve the above core problems
generalizes to novel 3D tasks (e.g., scene layout estimation, object pose
estimation, surface normal estimation) without the need for fine-tuning and
shows traits of abstraction abilities (e.g., cross-modality pose estimation).
In the context of the core supervised tasks, we demonstrate our representation
achieves state-of-the-art wide baseline feature matching results without
requiring apriori rectification (unlike SIFT and the majority of learned
features). We also show 6DOF camera pose estimation given a pair local image
patches. The accuracy of both supervised tasks come comparable to humans.
Finally, we contribute a large-scale dataset composed of object-centric street
view scenes along with point correspondences and camera pose information, and
conclude with a discussion on the learned representation and open research
questions.
|
Deep Learning applied to Road Traffic Speed forecasting | In this paper, we propose deep learning architectures (FNN, CNN and LSTM) to
forecast a regression model for time dependent data. These algorithm's are
designed to handle Floating Car Data (FCD) historic speeds to predict road
traffic data. For this we aggregate the speeds into the network inputs in an
innovative way. We compare the RMSE thus obtained with the results of a simpler
physical model, and show that the latter achieves better RMSE accuracy. We also
propose a new indicator, which evaluates the algorithms improvement when
compared to a benchmark prediction. We conclude by questioning the interest of
using deep learning methods for this specific regression task.
|
CollabLoc: Privacy-Preserving Multi-Modal Localization via Collaborative
Information Fusion | Mobile phones provide an excellent opportunity for building context-aware
applications. In particular, location-based services are important
context-aware services that are more and more used for enforcing security
policies, for supporting indoor room navigation, and for providing personalized
assistance. However, a major problem still remains unaddressed---the lack of
solutions that work across buildings while not using additional infrastructure
and also accounting for privacy and reliability needs. In this paper, a
privacy-preserving, multi-modal, cross-building, collaborative localization
platform is proposed based on Wi-Fi RSSI (existing infrastructure), Cellular
RSSI, sound and light levels, that enables room-level localization as main
application (though sub room level granularity is possible). The privacy is
inherently built into the solution based on onion routing, and
perturbation/randomization techniques, and exploits the idea of weighted
collaboration to increase the reliability as well as to limit the effect of
noisy devices (due to sensor noise/privacy). The proposed solution has been
analyzed in terms of privacy, accuracy, optimum parameters, and other overheads
on location data collected at multiple indoor and outdoor locations using an
Android app.
|
AutoEncoder Inspired Unsupervised Feature Selection | High-dimensional data in many areas such as computer vision and machine
learning tasks brings in computational and analytical difficulty. Feature
selection which selects a subset from observed features is a widely used
approach for improving performance and effectiveness of machine learning models
with high-dimensional data. In this paper, we propose a novel AutoEncoder
Feature Selector (AEFS) for unsupervised feature selection which combines
autoencoder regression and group lasso tasks. Compared to traditional feature
selection methods, AEFS can select the most important features by excavating
both linear and nonlinear information among features, which is more flexible
than the conventional self-representation method for unsupervised feature
selection with only linear assumptions. Experimental results on benchmark
dataset show that the proposed method is superior to the state-of-the-art
method.
|
Constrained Bayesian Active Learning of Interference Channels in
Cognitive Radio Networks | In this paper, a sequential probing method for interference constraint
learning is proposed to allow a centralized Cognitive Radio Network (CRN)
accessing the frequency band of a Primary User (PU) in an underlay cognitive
scenario with a designed PU protection specification. The main idea is that the
CRN probes the PU and subsequently eavesdrops the reverse PU link to acquire
the binary ACK/NACK packet. This feedback indicates whether the probing-induced
interference is harmful or not and can be used to learn the PU interference
constraint. The cognitive part of this sequential probing process is the
selection of the power levels of the Secondary Users (SUs) which aims to learn
the PU interference constraint with a minimum number of probing attempts while
setting a limit on the number of harmful probing-induced interference events or
equivalently of NACK packet observations over a time window. This constrained
design problem is studied within the Active Learning (AL) framework and an
optimal solution is derived and implemented with a sophisticated, accurate and
fast Bayesian Learning method, the Expectation Propagation (EP). The
performance of this solution is also demonstrated through numerical simulations
and compared with modified versions of AL techniques we developed in earlier
work.
|
Improving One-Shot Learning through Fusing Side Information | Deep Neural Networks (DNNs) often struggle with one-shot learning where we
have only one or a few labeled training examples per category. In this paper,
we argue that by using side information, we may compensate the missing
information across classes. We introduce two statistical approaches for fusing
side information into data representation learning to improve one-shot
learning. First, we propose to enforce the statistical dependency between data
representations and multiple types of side information. Second, we introduce an
attention mechanism to efficiently treat examples belonging to the
'lots-of-examples' classes as quasi-samples (additional training samples) for
'one-example' classes. We empirically show that our learning architecture
improves over traditional softmax regression networks as well as
state-of-the-art attentional regression networks on one-shot recognition tasks.
|
Stability and Generalization of Learning Algorithms that Converge to
Global Optima | We establish novel generalization bounds for learning algorithms that
converge to global minima. We do so by deriving black-box stability results
that only depend on the convergence of a learning algorithm and the geometry
around the minimizers of the loss function. The results are shown for nonconvex
loss functions satisfying the Polyak-{\L}ojasiewicz (PL) and the quadratic
growth (QG) conditions. We further show that these conditions arise for some
neural networks with linear activations. We use our black-box results to
establish the stability of optimization algorithms such as stochastic gradient
descent (SGD), gradient descent (GD), randomized coordinate descent (RCD), and
the stochastic variance reduced gradient method (SVRG), in both the PL and the
strongly convex setting. Our results match or improve state-of-the-art
generalization bounds and can easily be extended to similar optimization
algorithms. Finally, we show that although our results imply comparable
stability for SGD and GD in the PL setting, there exist simple neural networks
with multiple local minima where SGD is stable but GD is not.
|
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At
Every Step | Generative adversarial networks (GANs) are a family of generative models that
do not minimize a single training criterion. Unlike other generative models,
the data distribution is learned via a game between a generator (the generative
model) and a discriminator (a teacher providing training signal) that each
minimize their own cost. GANs are designed to reach a Nash equilibrium at which
each player cannot reduce their cost without changing the other players'
parameters. One useful approach for the theory of GANs is to show that a
divergence between the training distribution and the model distribution obtains
its minimum value at equilibrium. Several recent research directions have been
motivated by the idea that this divergence is the primary guide for the
learning process and that every step of learning should decrease the
divergence. We show that this view is overly restrictive. During GAN training,
the discriminator provides learning signal in situations where the gradients of
the divergences between distributions would not be useful. We provide empirical
counterexamples to the view of GAN training as divergence minimization.
Specifically, we demonstrate that GANs are able to learn distributions in
situations where the divergence minimization point of view predicts they would
fail. We also show that gradient penalties motivated from the divergence
minimization perspective are equally helpful when applied in other contexts in
which the divergence minimization perspective does not predict they would be
helpful. This contributes to a growing body of evidence that GAN training may
be more usefully viewed as approaching Nash equilibria via trajectories that do
not necessarily minimize a specific divergence at each step.
|
Interpretable Machine Learning for Privacy-Preserving Pervasive Systems | Our everyday interactions with pervasive systems generate traces that capture
various aspects of human behavior and enable machine learning algorithms to
extract latent information about users. In this paper, we propose a machine
learning interpretability framework that enables users to understand how these
generated traces violate their privacy.
|
A Unified Framework for Long Range and Cold Start Forecasting of
Seasonal Profiles in Time Series | Providing long-range forecasts is a fundamental challenge in time series
modeling, which is only compounded by the challenge of having to form such
forecasts when a time series has never previously been observed. The latter
challenge is the time series version of the cold-start problem seen in
recommender systems which, to our knowledge, has not been addressed in previous
work. A similar problem occurs when a long range forecast is required after
only observing a small number of time points --- a warm start forecast. With
these aims in mind, we focus on forecasting seasonal profiles---or baseline
demand---for periods on the order of a year in three cases: the long range case
with multiple previously observed seasonal profiles, the cold start case with
no previous observed seasonal profiles, and the warm start case with only a
single partially observed profile. Classical time series approaches that
perform iterated step-ahead forecasts based on previous observations struggle
to provide accurate long range predictions; in settings with little to no
observed data, such approaches are simply not applicable. Instead, we present a
straightforward framework which combines ideas from high-dimensional regression
and matrix factorization on a carefully constructed data matrix. Key to our
formulation and resulting performance is leveraging (1) repeated patterns over
fixed periods of time and across series, and (2) metadata associated with the
individual series; without this additional data, the cold-start/warm-start
problems are nearly impossible to solve. We demonstrate that our framework can
accurately forecast an array of seasonal profiles on multiple large scale
datasets.
|
Nesterov's Acceleration For Approximate Newton | Optimization plays a key role in machine learning. Recently, stochastic
second-order methods have attracted much attention due to their low
computational cost in each iteration. However, these algorithms might perform
poorly especially if it is hard to approximate the Hessian well and
efficiently. As far as we know, there is no effective way to handle this
problem. In this paper, we resort to Nesterov's acceleration technique to
improve the convergence performance of a class of second-order methods called
approximate Newton. We give a theoretical analysis that Nesterov's acceleration
technique can improve the convergence performance for approximate Newton just
like for first-order methods. We accordingly propose an accelerated regularized
sub-sampled Newton. Our accelerated algorithm performs much better than the
original regularized sub-sampled Newton in experiments, which validates our
theory empirically. Besides, the accelerated regularized sub-sampled Newton has
good performance comparable to or even better than classical algorithms.
|
Convolutional Neural Knowledge Graph Learning | Previous models for learning entity and relationship embeddings of knowledge
graphs such as TransE, TransH, and TransR aim to explore new links based on
learned representations. However, these models interpret relationships as
simple translations on entity embeddings. In this paper, we try to learn more
complex connections between entities and relationships. In particular, we use a
Convolutional Neural Network (CNN) to learn entity and relationship
representations in knowledge graphs. In our model, we treat entities and
relationships as one-dimensional numerical sequences with the same length.
After that, we combine each triplet of head, relationship, and tail together as
a matrix with height 3. CNN is applied to the triplets to get confidence
scores. Positive and manually corrupted negative triplets are used to train the
embeddings and the CNN model simultaneously. Experimental results on public
benchmark datasets show that the proposed model outperforms state-of-the-art
models on exploring unseen relationships, which proves that CNN is effective to
learn complex interactive patterns between entities and relationships.
|
Stability Analysis of Optimal Adaptive Control using Value Iteration
with Approximation Errors | Adaptive optimal control using value iteration initiated from a stabilizing
control policy is theoretically analyzed in terms of stability of the system
during the learning stage without ignoring the effects of approximation errors.
This analysis includes the system operated using any single/constant resulting
control policy and also using an evolving/time-varying control policy. A
feature of the presented results is providing estimations of the \textit{region
of attraction} so that if the initial condition is within the region, the whole
trajectory will remain inside it and hence, the function approximation results
remain valid.
|
Benchmark of Deep Learning Models on Large Healthcare MIMIC Datasets | Deep learning models (aka Deep Neural Networks) have revolutionized many
fields including computer vision, natural language processing, speech
recognition, and is being increasingly used in clinical healthcare
applications. However, few works exist which have benchmarked the performance
of the deep learning models with respect to the state-of-the-art machine
learning models and prognostic scoring systems on publicly available healthcare
datasets. In this paper, we present the benchmarking results for several
clinical prediction tasks such as mortality prediction, length of stay
prediction, and ICD-9 code group prediction using Deep Learning models,
ensemble of machine learning models (Super Learner algorithm), SAPS II and SOFA
scores. We used the Medical Information Mart for Intensive Care III (MIMIC-III)
(v1.4) publicly available dataset, which includes all patients admitted to an
ICU at the Beth Israel Deaconess Medical Center from 2001 to 2012, for the
benchmarking tasks. Our results show that deep learning models consistently
outperform all the other approaches especially when the `raw' clinical time
series data is used as input features to the models.
|
Neural Stain-Style Transfer Learning using GAN for Histopathological
Images | Performance of data-driven network for tumor classification varies with
stain-style of histopathological images. This article proposes the stain-style
transfer (SST) model based on conditional generative adversarial networks
(GANs) which is to learn not only the certain color distribution but also the
corresponding histopathological pattern. Our model considers feature-preserving
loss in addition to well-known GAN loss. Consequently our model does not only
transfers initial stain-styles to the desired one but also prevent the
degradation of tumor classifier on transferred images. The model is examined
using the CAMELYON16 dataset.
|
Max-Margin Invariant Features from Transformed Unlabeled Data | The study of representations invariant to common transformations of the data
is important to learning. Most techniques have focused on local approximate
invariance implemented within expensive optimization frameworks lacking
explicit theoretical guarantees. In this paper, we study kernels that are
invariant to a unitary group while having theoretical guarantees in addressing
the important practical issue of unavailability of transformed versions of
labelled data. A problem we call the Unlabeled Transformation Problem which is
a special form of semi-supervised learning and one-shot learning. We present a
theoretically motivated alternate approach to the invariant kernel SVM based on
which we propose Max-Margin Invariant Features (MMIF) to solve this problem. As
an illustration, we design an framework for face recognition and demonstrate
the efficacy of our approach on a large scale semi-synthetic dataset with
153,000 images and a new challenging protocol on Labelled Faces in the Wild
(LFW) while out-performing strong baselines.
|
Interpretable Deep Learning applied to Plant Stress Phenotyping | Availability of an explainable deep learning model that can be applied to
practical real world scenarios and in turn, can consistently, rapidly and
accurately identify specific and minute traits in applicable fields of
biological sciences, is scarce. Here we consider one such real world example
viz., accurate identification, classification and quantification of biotic and
abiotic stresses in crop research and production. Up until now, this has been
predominantly done manually by visual inspection and require specialized
training. However, such techniques are hindered by subjectivity resulting from
inter- and intra-rater cognitive variability. Here, we demonstrate the ability
of a machine learning framework to identify and classify a diverse set of
foliar stresses in the soybean plant with remarkable accuracy. We also present
an explanation mechanism using gradient-weighted class activation mapping that
isolates the visual symptoms used by the model to make predictions. This
unsupervised identification of unique visual symptoms for each stress provides
a quantitative measure of stress severity, allowing for identification,
classification and quantification in one framework. The learnt model appears to
be agnostic to species and make good predictions for other (non-soybean)
species, demonstrating an ability of transfer learning.
|
Improving Accuracy of Nonparametric Transfer Learning via Vector
Segmentation | Transfer learning using deep neural networks as feature extractors has become
increasingly popular over the past few years. It allows to obtain
state-of-the-art accuracy on datasets too small to train a deep neural network
on its own, and it provides cutting edge descriptors that, combined with
nonparametric learning methods, allow rapid and flexible deployment of
performing solutions in computationally restricted settings. In this paper, we
are interested in showing that the features extracted using deep neural
networks have specific properties which can be used to improve accuracy of
downstream nonparametric learning methods. Namely, we demonstrate that for some
distributions where information is embedded in a few coordinates, segmenting
feature vectors can lead to better accuracy. We show how this model can be
applied to real datasets by performing experiments using three mainstream deep
neural network feature extractors and four databases, in vision and audio.
|
Auto-Differentiating Linear Algebra | Development systems for deep learning (DL), such as Theano, Torch,
TensorFlow, or MXNet, are easy-to-use tools for creating complex neural network
models. Since gradient computations are automatically baked in, and execution
is mapped to high performance hardware, these models can be trained end-to-end
on large amounts of data. However, it is currently not easy to implement many
basic machine learning primitives in these systems (such as Gaussian processes,
least squares estimation, principal components analysis, Kalman smoothing),
mainly because they lack efficient support of linear algebra primitives as
differentiable operators. We detail how a number of matrix decompositions
(Cholesky, LQ, symmetric eigen) can be implemented as differentiable operators.
We have implemented these primitives in MXNet, running on CPU and GPU in single
and double precision. We sketch use cases of these new operators, learning
Gaussian process and Bayesian linear regression models, where we demonstrate
very substantial reductions in implementation complexity and running time
compared to previous codes. Our MXNet extension allows end-to-end learning of
hybrid models, which combine deep neural networks (DNNs) with Bayesian
concepts, with applications in advanced Gaussian process models, scalable
Bayesian optimization, and Bayesian active learning.
|
A Correction Method of a Binary Classifier Applied to Multi-label
Pairwise Models | In this work, we addressed the issue of applying a stochastic classifier and
a local, fuzzy confusion matrix under the framework of multi-label
classification. We proposed a novel solution to the problem of correcting label
pairwise ensembles. The main step of the correction procedure is to compute
classifier- specific competence and cross-competence measures, which estimates
error pattern of the underlying classifier. We considered two improvements of
the method of obtaining confusion matrices. The first one is aimed to deal with
imbalanced labels. The other utilizes double labelled instances which are
usually removed during the pairwise transformation. The proposed methods were
evaluated using 29 benchmark datasets. In order to assess the efficiency of the
introduced models, they were compared against 1 state-of-the-art approach and
the correction scheme based on the original method of confusion matrix
estimation. The comparison was performed using four different multi-label
evaluation measures: macro and micro-averaged F1 loss, zero-one loss and
Hamming loss. Additionally, we investigated relations between classification
quality, which is expressed in terms of different quality criteria, and
characteristics of multi-label datasets such as average imbalance ratio or
label density. The experimental study reveals that the correction approaches
significantly outperforms the reference method only in terms of zero-one loss.
|
A Bayesian Method for Joint Clustering of Vectorial Data and Network
Data | We present a new model-based integrative method for clustering objects given
both vectorial data, which describes the feature of each object, and network
data, which indicates the similarity of connected objects. The proposed general
model is able to cluster the two types of data simultaneously within one
integrative probabilistic model, while traditional methods can only handle one
data type or depend on transforming one data type to another. Bayesian
inference of the clustering is conducted based on a Markov chain Monte Carlo
algorithm. A special case of the general model combining the Gaussian mixture
model and the stochastic block model is extensively studied. We used both
synthetic data and real data to evaluate this new method and compare it with
alternative methods. The results show that our simultaneous clustering method
performs much better. This improvement is due to the power of the model-based
probabilistic approach for efficiently integrating information.
|
One pixel attack for fooling deep neural networks | Recent research has revealed that the output of Deep Neural Networks (DNN)
can be easily altered by adding relatively small perturbations to the input
vector. In this paper, we analyze an attack in an extremely limited scenario
where only one pixel can be modified. For that we propose a novel method for
generating one-pixel adversarial perturbations based on differential evolution
(DE). It requires less adversarial information (a black-box attack) and can
fool more types of networks due to the inherent features of DE. The results
show that 67.97% of the natural images in Kaggle CIFAR-10 test dataset and
16.04% of the ImageNet (ILSVRC 2012) test images can be perturbed to at least
one target class by modifying just one pixel with 74.03% and 22.91% confidence
on average. We also show the same vulnerability on the original CIFAR-10
dataset. Thus, the proposed attack explores a different take on adversarial
machine learning in an extreme limited scenario, showing that current DNNs are
also vulnerable to such low dimension attacks. Besides, we also illustrate an
important application of DE (or broadly speaking, evolutionary computation) in
the domain of adversarial machine learning: creating tools that can effectively
generate low-cost adversarial attacks against neural networks for evaluating
robustness.
|
Classification on Large Networks: A Quantitative Bound via Motifs and
Graphons | When each data point is a large graph, graph statistics such as densities of
certain subgraphs (motifs) can be used as feature vectors for machine learning.
While intuitive, motif counts are expensive to compute and difficult to work
with theoretically. Via graphon theory, we give an explicit quantitative bound
for the ability of motif homomorphisms to distinguish large networks under both
generative and sampling noise. Furthermore, we give similar bounds for the
graph spectrum and connect it to homomorphism densities of cycles. This results
in an easily computable classifier on graph data with theoretical performance
guarantee. Our method yields competitive results on classification tasks for
the autoimmune disease Lupus Erythematosus.
|
Avoiding Communication in Proximal Methods for Convex Optimization
Problems | The fast iterative soft thresholding algorithm (FISTA) is used to solve
convex regularized optimization problems in machine learning. Distributed
implementations of the algorithm have become popular since they enable the
analysis of large datasets. However, existing formulations of FISTA communicate
data at every iteration which reduces its performance on modern distributed
architectures. The communication costs of FISTA, including bandwidth and
latency costs, is closely tied to the mathematical formulation of the
algorithm. This work reformulates FISTA to communicate data at every k
iterations and reduce data communication when operating on large data sets. We
formulate the algorithm for two different optimization methods on the Lasso
problem and show that the latency cost is reduced by a factor of k while
bandwidth and floating-point operation costs remain the same. The convergence
rates and stability properties of the reformulated algorithms are similar to
the standard formulations. The performance of communication-avoiding FISTA and
Proximal Newton methods is evaluated on 1 to 1024 nodes for multiple benchmarks
and demonstrate average speedups of 3-10x with scaling properties that
outperform the classical algorithms.
|
Fast Model Identification via Physics Engines for Data-Efficient Policy
Search | This paper presents a method for identifying mechanical parameters of robots
or objects, such as their mass and friction coefficients. Key features are the
use of off-the-shelf physics engines and the adaptation of a Bayesian
optimization technique towards minimizing the number of real-world experiments
needed for model-based reinforcement learning. The proposed framework
reproduces in a physics engine experiments performed on a real robot and
optimizes the model's mechanical parameters so as to match real-world
trajectories. The optimized model is then used for learning a policy in
simulation, before real-world deployment. It is well understood, however, that
it is hard to exactly reproduce real trajectories in simulation. Moreover, a
near-optimal policy can be frequently found with an imperfect model. Therefore,
this work proposes a strategy for identifying a model that is just good enough
to approximate the value of a locally optimal policy with a certain confidence,
instead of wasting effort on identifying the most accurate model. Evaluations,
performed both in simulation and on a real robotic manipulation task, indicate
that the proposed strategy results in an overall time-efficient, integrated
model identification and learning solution, which significantly improves the
data-efficiency of existing policy search algorithms.
|
Conformal predictive distributions with kernels | This paper reviews the checkered history of predictive distributions in
statistics and discusses two developments, one from recent literature and the
other new. The first development is bringing predictive distributions into
machine learning, whose early development was so deeply influenced by two
remarkable groups at the Institute of Automation and Remote Control. The second
development is combining predictive distributions with kernel methods, which
were originated by one of those groups, including Emmanuel Braverman.
|
Curvature-aided Incremental Aggregated Gradient Method | We propose a new algorithm for finite sum optimization which we call the
curvature-aided incremental aggregated gradient (CIAG) method. Motivated by the
problem of training a classifier for a d-dimensional problem, where the number
of training data is $m$ and $m \gg d \gg 1$, the CIAG method seeks to
accelerate incremental aggregated gradient (IAG) methods using aids from the
curvature (or Hessian) information, while avoiding the evaluation of matrix
inverses required by the incremental Newton (IN) method. Specifically, our idea
is to exploit the incrementally aggregated Hessian matrix to trace the full
gradient vector at every incremental step, therefore achieving an improved
linear convergence rate over the state-of-the-art IAG methods. For strongly
convex problems, the fast linear convergence rate requires the objective
function to be close to quadratic, or the initial point to be close to optimal
solution. Importantly, we show that running one iteration of the CIAG method
yields the same improvement to the optimality gap as running one iteration of
the full gradient method, while the complexity is $O(d^2)$ for CIAG and $O(md)$
for the full gradient. Overall, the CIAG method strikes a balance between the
high computation complexity incremental Newton-type methods and the slow IAG
method. Our numerical results support the theoretical findings and show that
the CIAG method often converges with much fewer iterations than IAG, and
requires much shorter running time than IN when the problem dimension is high.
|
Exact Mean Computation in Dynamic Time Warping Spaces | Dynamic time warping constitutes a major tool for analyzing time series. In
particular, computing a mean series of a given sample of series in dynamic time
warping spaces (by minimizing the Fr\'echet function) is a challenging
computational problem, so far solved by several heuristic and inexact
strategies. We spot some inaccuracies in the literature on exact mean
computation in dynamic time warping spaces. Our contributions comprise an exact
dynamic program computing a mean (useful for benchmarking and evaluating known
heuristics). Based on this dynamic program, we empirically study properties
like uniqueness and length of a mean. Moreover, experimental evaluations reveal
substantial deficits of state-of-the-art heuristics in terms of their output
quality. We also give an exact polynomial-time algorithm for the special case
of binary time series.
|
Fast and Scalable Distributed Deep Convolutional Autoencoder for fMRI
Big Data Analytics | In recent years, analyzing task-based fMRI (tfMRI) data has become an
essential tool for understanding brain function and networks. However, due to
the sheer size of tfMRI data, its intrinsic complex structure, and lack of
ground truth of underlying neural activities, modeling tfMRI data is hard and
challenging. Previously proposed data-modeling methods including Independent
Component Analysis (ICA) and Sparse Dictionary Learning only provided a weakly
established model based on blind source separation under the strong assumption
that original fMRI signals could be linearly decomposed into time series
components with corresponding spatial maps. Meanwhile, analyzing and learning a
large amount of tfMRI data from a variety of subjects has been shown to be very
demanding but yet challenging even with technological advances in computational
hardware. Given the Convolutional Neural Network (CNN), a robust method for
learning high-level abstractions from low-level data such as tfMRI time series,
in this work we propose a fast and scalable novel framework for distributed
deep Convolutional Autoencoder model. This model aims to both learn the complex
hierarchical structure of the tfMRI data and to leverage the processing power
of multiple GPUs in a distributed fashion. To implement such a model, we have
created an enhanced processing pipeline on the top of Apache Spark and
Tensorflow library, leveraging from a very large cluster of GPU machines.
Experimental data from applying the model on the Human Connectome Project (HCP)
show that the proposed model is efficient and scalable toward tfMRI big data
analytics, thus enabling data-driven extraction of hierarchical neuroscientific
information from massive fMRI big data in the future.
|
Scaling Text with the Class Affinity Model | Probabilistic methods for classifying text form a rich tradition in machine
learning and natural language processing. For many important problems, however,
class prediction is uninteresting because the class is known, and instead the
focus shifts to estimating latent quantities related to the text, such as
affect or ideology. We focus on one such problem of interest, estimating the
ideological positions of 55 Irish legislators in the 1991 D\'ail confidence
vote. To solve the D\'ail scaling problem and others like it, we develop a text
modeling framework that allows actors to take latent positions on a "gray"
spectrum between "black" and "white" polar opposites. We are able to validate
results from this model by measuring the influences exhibited by individual
words, and we are able to quantify the uncertainty in the scaling estimates by
using a sentence-level block bootstrap. Applying our method to the D\'ail
debate, we are able to scale the legislators between extreme pro-government and
pro-opposition in a way that reveals nuances in their speeches not captured by
their votes or party affiliations.
|
Efficiently Trainable Text-to-Speech System Based on Deep Convolutional
Networks with Guided Attention | This paper describes a novel text-to-speech (TTS) technique based on deep
convolutional neural networks (CNN), without use of any recurrent units.
Recurrent neural networks (RNN) have become a standard technique to model
sequential data recently, and this technique has been used in some cutting-edge
neural TTS techniques. However, training RNN components often requires a very
powerful computer, or a very long time, typically several days or weeks. Recent
other studies, on the other hand, have shown that CNN-based sequence synthesis
can be much faster than RNN-based techniques, because of high
parallelizability. The objective of this paper is to show that an alternative
neural TTS based only on CNN alleviate these economic costs of training. In our
experiment, the proposed Deep Convolutional TTS was sufficiently trained
overnight (15 hours), using an ordinary gaming PC equipped with two GPUs, while
the quality of the synthesized speech was almost acceptable.
|
Multi-Armed Bandits with Metric Movement Costs | We consider the non-stochastic Multi-Armed Bandit problem in a setting where
there is a fixed and known metric on the action space that determines a cost
for switching between any pair of actions. The loss of the online learner has
two components: the first is the usual loss of the selected actions, and the
second is an additional loss due to switching between actions. Our main
contribution gives a tight characterization of the expected minimax regret in
this setting, in terms of a complexity measure $\mathcal{C}$ of the underlying
metric which depends on its covering numbers. In finite metric spaces with $k$
actions, we give an efficient algorithm that achieves regret of the form
$\widetilde{O}(\max\{\mathcal{C}^{1/3}T^{2/3},\sqrt{kT}\})$, and show that this
is the best possible. Our regret bound generalizes previous known regret bounds
for some special cases: (i) the unit-switching cost regret
$\widetilde{\Theta}(\max\{k^{1/3}T^{2/3},\sqrt{kT}\})$ where
$\mathcal{C}=\Theta(k)$, and (ii) the interval metric with regret
$\widetilde{\Theta}(\max\{T^{2/3},\sqrt{kT}\})$ where $\mathcal{C}=\Theta(1)$.
For infinite metrics spaces with Lipschitz loss functions, we derive a tight
regret bound of $\widetilde{\Theta}(T^{\frac{d+1}{d+2}})$ where $d \ge 1$ is
the Minkowski dimension of the space, which is known to be tight even when
there are no switching costs.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.