title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Attributes' Importance for Zero-Shot Pose-Classification Based on
Wearable Sensors
|
This paper presents a simple yet effective method for improving the
performance of zero-shot learning (ZSL). ZSL classifies instances of unseen
classes, from which no training data is available, by utilizing the attributes
of the classes. Conventional ZSL methods have equally dealt with all the
available attributes, but this sometimes causes misclassification. This is
because an attribute that is effective for classifying instances of one class
is not always effective for another class. In this case, a metric of
classifying the latter class can be undesirably influenced by the irrelevant
attribute. This paper solves this problem by taking the importance of each
attribute for each class into account when calculating the metric. In addition
to the proposal of this new method, this paper also contributes by providing a
dataset for pose classification based on wearable sensors, named HDPoseDS. It
contains 22 classes of poses performed by 10 subjects with 31 IMU sensors
across full body. To the best of our knowledge, it is the richest
wearable-sensor dataset especially in terms of sensor density, and thus it is
suitable for studying zero-shot pose/action recognition. The presented method
was evaluated on HDPoseDS and outperformed relative improvement of 5.9% in
comparison to the best baseline method.
|
Large Scale Language Modeling: Converging on 40GB of Text in Four Hours
|
Recent work has shown how to train Convolutional Neural Networks (CNNs)
rapidly on large image datasets, then transfer the knowledge gained from these
models to a variety of tasks. Following [Radford 2017], in this work, we
demonstrate similar scalability and transfer for Recurrent Neural Networks
(RNNs) for Natural Language tasks. By utilizing mixed precision arithmetic and
a 32k batch size distributed across 128 NVIDIA Tesla V100 GPUs, we are able to
train a character-level 4096-dimension multiplicative LSTM (mLSTM) for
unsupervised text reconstruction over 3 epochs of the 40 GB Amazon Reviews
dataset in four hours. This runtime compares favorably with previous work
taking one month to train the same size and configuration for one epoch over
the same dataset. Converging large batch RNN models can be challenging. Recent
work has suggested scaling the learning rate as a function of batch size, but
we find that simply scaling the learning rate as a function of batch size leads
either to significantly worse convergence or immediate divergence for this
problem. We provide a learning rate schedule that allows our model to converge
with a 32k batch size. Since our model converges over the Amazon Reviews
dataset in hours, and our compute requirement of 128 Tesla V100 GPUs, while
substantial, is commercially available, this work opens up large scale
unsupervised NLP training to most commercial applications and deep learning
researchers. A model can be trained over most public or private text datasets
overnight.
|
code2seq: Generating Sequences from Structured Representations of Code
|
The ability to generate natural language sequences from source code snippets
has a variety of applications such as code summarization, documentation, and
retrieval. Sequence-to-sequence (seq2seq) models, adopted from neural machine
translation (NMT), have achieved state-of-the-art performance on these tasks by
treating source code as a sequence of tokens. We present ${\rm {\scriptsize
CODE2SEQ}}$: an alternative approach that leverages the syntactic structure of
programming languages to better encode source code. Our model represents a code
snippet as the set of compositional paths in its abstract syntax tree (AST) and
uses attention to select the relevant paths while decoding. We demonstrate the
effectiveness of our approach for two tasks, two programming languages, and
four datasets of up to $16$M examples. Our model significantly outperforms
previous models that were specifically designed for programming languages, as
well as state-of-the-art NMT models. An interactive online demo of our model is
available at http://code2seq.org. Our code, data and trained models are
available at http://github.com/tech-srl/code2seq.
|
Predicting Expressive Speaking Style From Text In End-To-End Speech
Synthesis
|
Global Style Tokens (GSTs) are a recently-proposed method to learn latent
disentangled representations of high-dimensional data. GSTs can be used within
Tacotron, a state-of-the-art end-to-end text-to-speech synthesis system, to
uncover expressive factors of variation in speaking style. In this work, we
introduce the Text-Predicted Global Style Token (TP-GST) architecture, which
treats GST combination weights or style embeddings as "virtual" speaking style
labels within Tacotron. TP-GST learns to predict stylistic renderings from text
alone, requiring neither explicit labels during training nor auxiliary inputs
for inference. We show that, when trained on a dataset of expressive speech,
our system generates audio with more pitch and energy variation than two
state-of-the-art baseline models. We further demonstrate that TP-GSTs can
synthesize speech with background noise removed, and corroborate these analyses
with positive results on human-rated listener preference audiobook tasks.
Finally, we demonstrate that multi-speaker TP-GST models successfully factorize
speaker identity and speaking style. We provide a website with audio samples
for each of our findings.
|
Spatial Deep Learning for Wireless Scheduling
|
The optimal scheduling of interfering links in a dense wireless network with
full frequency reuse is a challenging task. The traditional method involves
first estimating all the interfering channel strengths then optimizing the
scheduling based on the model. This model-based method is however resource
intensive and computationally hard because channel estimation is expensive in
dense networks; furthermore, finding even a locally optimal solution of the
resulting optimization problem may be computationally complex. This paper shows
that by using a deep learning approach, it is possible to bypass the channel
estimation and to schedule links efficiently based solely on the geographic
locations of the transmitters and the receivers, due to the fact that in many
propagation environments, the wireless channel strength is largely a function
of the distance dependent path-loss. This is accomplished by unsupervised
training over randomly deployed networks, and by using a novel neural network
architecture that computes the geographic spatial convolutions of the
interfering or interfered neighboring nodes along with subsequent multiple
feedback stages to learn the optimum solution. The resulting neural network
gives near-optimal performance for sum-rate maximization and is capable of
generalizing to larger deployment areas and to deployments of different link
densities. Moreover, to provide fairness, this paper proposes a novel
scheduling approach that utilizes the sum-rate optimal scheduling algorithm
over judiciously chosen subsets of links for maximizing a proportional fairness
objective over the network. The proposed approach shows highly competitive and
generalizable network utility maximization results.
|
DELIMIT PyTorch - An extension for Deep Learning in Diffusion Imaging
|
DELIMIT is a framework extension for deep learning in diffusion imaging,
which extends the basic framework PyTorch towards spherical signals. Based on
several novel layers, deep learning can be applied to spherical diffusion
imaging data in a very convenient way. First, two spherical harmonic
interpolation layers are added to the extension, which allow to transform the
signal from spherical surface space into the spherical harmonic space, and vice
versa. In addition, a local spherical convolution layer is introduced that adds
the possibility to include gradient neighborhood information within the
network. Furthermore, these extensions can also be utilized for the
preprocessing of diffusion signals.
|
Learning disentangled representation from 12-lead electrograms:
application in localizing the origin of Ventricular Tachycardia
|
The increasing availability of electrocardiogram (ECG) data has motivated the
use of data-driven models for automating various clinical tasks based on ECG
data. The development of subject-specific models are limited by the cost and
difficulty of obtaining sufficient training data for each individual. The
alternative of population model, however, faces challenges caused by the
significant inter-subject variations within the ECG data. We address this
challenge by investigating for the first time the problem of learning
representations for clinically-informative variables while disentangling other
factors of variations within the ECG data. In this work, we present a
conditional variational autoencoder (VAE) to extract the subject-specific
adjustment to the ECG data, conditioned on task-specific representations
learned from a deterministic encoder. To encourage the representation for
inter-subject variations to be independent from the task-specific
representation, maximum mean discrepancy is used to match all the moments
between the distributions learned by the VAE conditioning on the code from the
deterministic encoder. The learning of the task-specific representation is
regularized by a weak supervision in the form of contrastive regularization. We
apply the proposed method to a novel yet important clinical task of classifying
the origin of ventricular tachycardia (VT) into pre-defined segments,
demonstrating the efficacy of the proposed method against the standard VAE.
|
Deep Reinforcement One-Shot Learning for Artificially Intelligent
Classification Systems
|
In recent years there has been a sharp rise in networking applications, in
which significant events need to be classified but only a few training
instances are available. These are known as cases of one-shot learning.
Examples include analyzing network traffic under zero-day attacks, and computer
vision tasks by sensor networks deployed in the field. To handle this
challenging task, organizations often use human analysts to classify events
under high uncertainty. Existing algorithms use a threshold-based mechanism to
decide whether to classify an object automatically or send it to an analyst for
deeper inspection. However, this approach leads to a significant waste of
resources since it does not take the practical temporal constraints of system
resources into account. Our contribution is threefold. First, we develop a
novel Deep Reinforcement One-shot Learning (DeROL) framework to address this
challenge. The basic idea of the DeROL algorithm is to train a deep-Q network
to obtain a policy which is oblivious to the unseen classes in the testing
data. Then, in real-time, DeROL maps the current state of the one-shot learning
process to operational actions based on the trained deep-Q network, to maximize
the objective function. Second, we develop the first open-source software for
practical artificially intelligent one-shot classification systems with limited
resources for the benefit of researchers in related fields. Third, we present
an extensive experimental study using the OMNIGLOT dataset for computer vision
tasks and the UNSW-NB15 dataset for intrusion detection tasks that demonstrates
the versatility and efficiency of the DeROL framework.
|
Global Convergence to the Equilibrium of GANs using Variational
Inequalities
|
In optimization, the negative gradient of a function denotes the direction of
steepest descent. Furthermore, traveling in any direction orthogonal to the
gradient maintains the value of the function. In this work, we show that these
orthogonal directions that are ignored by gradient descent can be critical in
equilibrium problems. Equilibrium problems have drawn heightened attention in
machine learning due to the emergence of the Generative Adversarial Network
(GAN). We use the framework of Variational Inequalities to analyze popular
training algorithms for a fundamental GAN variant: the Wasserstein
Linear-Quadratic GAN. We show that the steepest descent direction causes
divergence from the equilibrium, and convergence to the equilibrium is achieved
through following a particular orthogonal direction. We call this successful
technique Crossing-the-Curl, named for its mathematical derivation as well as
its intuition: identify the game's axis of rotation and move "across" space in
the direction towards smaller "curling".
|
Triplet Network with Attention for Speaker Diarization
|
In automatic speech processing systems, speaker diarization is a crucial
front-end component to separate segments from different speakers. Inspired by
the recent success of deep neural networks (DNNs) in semantic inferencing,
triplet loss-based architectures have been successfully used for this problem.
However, existing work utilizes conventional i-vectors as the input
representation and builds simple fully connected networks for metric learning,
thus not fully leveraging the modeling power of DNN architectures. This paper
investigates the importance of learning effective representations from the
sequences directly in metric learning pipelines for speaker diarization. More
specifically, we propose to employ attention models to learn embeddings and the
metric jointly in an end-to-end fashion. Experiments are conducted on the
CALLHOME conversational speech corpus. The diarization results demonstrate
that, besides providing a unified model, the proposed approach achieves
improved performance when compared against existing approaches.
|
Designing Adaptive Neural Networks for Energy-Constrained Image
Classification
|
As convolutional neural networks (CNNs) enable state-of-the-art computer
vision applications, their high energy consumption has emerged as a key
impediment to their deployment on embedded and mobile devices. Towards
efficient image classification under hardware constraints, prior work has
proposed adaptive CNNs, i.e., systems of networks with different accuracy and
computation characteristics, where a selection scheme adaptively selects the
network to be evaluated for each input image. While previous efforts have
investigated different network selection schemes, we find that they do not
necessarily result in energy savings when deployed on mobile systems. The key
limitation of existing methods is that they learn only how data should be
processed among the CNNs and not the network architectures, with each network
being treated as a blackbox.
To address this limitation, we pursue a more powerful design paradigm where
the architecture settings of the CNNs are treated as hyper-parameters to be
globally optimized. We cast the design of adaptive CNNs as a hyper-parameter
optimization problem with respect to energy, accuracy, and communication
constraints imposed by the mobile device. To efficiently solve this problem, we
adapt Bayesian optimization to the properties of the design space, reaching
near-optimal configurations in few tens of function evaluations. Our method
reduces the energy consumed for image classification on a mobile device by up
to 6x, compared to the best previously published work that uses CNNs as
blackboxes. Finally, we evaluate two image classification practices, i.e.,
classifying all images locally versus over the cloud under energy and
communication constraints.
|
Smart City Development with Urban Transfer Learning
|
Nowadays, the smart city development levels of different cities are still
unbalanced. For a large number of cities which just started development, the
governments will face a critical cold-start problem: 'how to develop a new
smart city service with limited data?'. To address this problem, transfer
learning can be leveraged to accelerate the smart city development, which we
term the urban transfer learning paradigm. This article investigates the common
process of urban transfer learning, aiming to provide city planners and
relevant practitioners with guidelines on how to apply this novel learning
paradigm. Our guidelines include common transfer strategies to take, general
steps to follow, and case studies in public safety, transportation management,
etc. We also summarize a few research opportunities and expect this article can
attract more researchers to study urban transfer learning.
|
Stock Price Correlation Coefficient Prediction with ARIMA-LSTM Hybrid
Model
|
Predicting the price correlation of two assets for future time periods is
important in portfolio optimization. We apply LSTM recurrent neural networks
(RNN) in predicting the stock price correlation coefficient of two individual
stocks. RNNs are competent in understanding temporal dependencies. The use of
LSTM cells further enhances its long term predictive properties. To encompass
both linearity and nonlinearity in the model, we adopt the ARIMA model as well.
The ARIMA model filters linear tendencies in the data and passes on the
residual value to the LSTM model. The ARIMA LSTM hybrid model is tested against
other traditional predictive financial models such as the full historical
model, constant correlation model, single index model and the multi group
model. In our empirical study, the predictive ability of the ARIMA-LSTM model
turned out superior to all other financial models by a significant scale. Our
work implies that it is worth considering the ARIMA LSTM model to forecast
correlation coefficient for portfolio optimization.
|
Autoencoder Based Sample Selection for Self-Taught Learning
|
Self-taught learning is a technique that uses a large number of unlabeled
data as source samples to improve the task performance on target samples.
Compared with other transfer learning techniques, self-taught learning can be
applied to a broader set of scenarios due to the loose restrictions on the
source data. However, knowledge transferred from source samples that are not
sufficiently related to the target domain may negatively influence the target
learner, which is referred to as negative transfer. In this paper, we propose a
metric for the relevance between a source sample and the target samples. To be
more specific, both source and target samples are reconstructed through a
single-layer autoencoder with a linear relationship between source samples and
reconstructed target samples being simultaneously enforced. An
$\ell_{2,1}$-norm sparsity constraint is imposed on the transformation matrix
to identify source samples relevant to the target domain. Source domain samples
that are deemed relevant are assigned pseudo-labels reflecting their relevance
to target domain samples, and are combined with target samples in order to
provide an expanded training set for classifier training. Local data structures
are also preserved during source sample selection through spectral graph
analysis. Promising results in extensive experiments show the advantages of the
proposed approach.
|
LISA: Explaining Recurrent Neural Network Judgments via Layer-wIse
Semantic Accumulation and Example to Pattern Transformation
|
Recurrent neural networks (RNNs) are temporal networks and cumulative in
nature that have shown promising results in various natural language processing
tasks. Despite their success, it still remains a challenge to understand their
hidden behavior. In this work, we analyze and interpret the cumulative nature
of RNN via a proposed technique named as Layer-wIse-Semantic-Accumulation
(LISA) for explaining decisions and detecting the most likely (i.e., saliency)
patterns that the network relies on while decision making. We demonstrate (1)
LISA: "How an RNN accumulates or builds semantics during its sequential
processing for a given text example and expected response" (2) Example2pattern:
"How the saliency patterns look like for each category in the data according to
the network in decision making". We analyse the sensitiveness of RNNs about
different inputs to check the increase or decrease in prediction scores and
further extract the saliency patterns learned by the network. We employ two
relation classification datasets: SemEval 10 Task 8 and TAC KBP Slot Filling to
explain RNN predictions via the LISA and example2pattern.
|
Spherical Harmonic Residual Network for Diffusion Signal Harmonization
|
Diffusion imaging is an important method in the field of neuroscience, as it
is sensitive to changes within the tissue microstructure of the human brain.
However, a major challenge when using MRI to derive quantitative measures is
that the use of different scanners, as used in multi-site group studies,
introduces measurement variability. This can lead to an increased variance in
quantitative metrics, even if the same brain is scanned.
Contrary to the assumption that these characteristics are comparable and
similar, small changes in these values are observed in many clinical studies,
hence harmonization of the signals is essential.
In this paper, we present a method that does not require additional
preprocessing, such as segmentation or registration, and harmonizes the signal
based on a deep learning residual network. For this purpose, a training
database is required, which consist of the same subjects, scanned on different
scanners.
The results show that harmonized signals are significantly more similar to
the ground truth signal compared to no harmonization, but also improve in
comparison to another deep learning method. The same effect is also
demonstrated in commonly used metrics derived from the diffusion MRI signal.
|
Using Machine Learning Safely in Automotive Software: An Assessment and
Adaption of Software Process Requirements in ISO 26262
|
The use of machine learning (ML) is on the rise in many sectors of software
development, and automotive software development is no different. In
particular, Advanced Driver Assistance Systems (ADAS) and Automated Driving
Systems (ADS) are two areas where ML plays a significant role. In automotive
development, safety is a critical objective, and the emergence of standards
such as ISO 26262 has helped focus industry practices to address safety in a
systematic and consistent way. Unfortunately, these standards were not designed
to accommodate technologies such as ML or the type of functionality that is
provided by an ADS and this has created a conflict between the need to innovate
and the need to improve safety. In this report, we take steps to address this
conflict by doing a detailed assessment and adaption of ISO 26262 for ML,
specifically in the context of supervised learning. First we analyze the key
factors that are the source of the conflict. Then we assess each software
development process requirement (Part 6 of ISO 26262) for applicability to ML.
Where there are gaps, we propose new requirements to address the gaps. Finally
we discuss the application of this adapted and extended variant of Part 6 to ML
development scenarios.
|
Predicting Learning Status in MOOCs using LSTM
|
Real-time and open online course resources of MOOCs have attracted a large
number of learners in recent years. However, many new questions were emerging
about the high dropout rate of learners. For MOOCs platform, predicting the
learning status of MOOCs learners in real time with high accuracy is the
crucial task, and it also help improve the quality of MOOCs teaching. The
prediction task in this paper is inherently a time series prediction problem,
and can be treated as time series classification problem, hence this paper
proposed a prediction model based on RNNLSTMs and optimization techniques which
can be used to predict learners' learning status. Using datasets provided by
Chinese University MOOCs as the inputs of model, the average accuracy of
model's outputs was about 90%.
|
A Review of Learning with Deep Generative Models from Perspective of
Graphical Modeling
|
This document aims to provide a review on learning with deep generative
models (DGMs), which is an highly-active area in machine learning and more
generally, artificial intelligence. This review is not meant to be a tutorial,
but when necessary, we provide self-contained derivations for completeness.
This review has two features. First, though there are different perspectives to
classify DGMs, we choose to organize this review from the perspective of
graphical modeling, because the learning methods for directed DGMs and
undirected DGMs are fundamentally different. Second, we differentiate model
definitions from model learning algorithms, since different learning algorithms
can be applied to solve the learning problem on the same model, and an
algorithm can be applied to learn different models. We thus separate model
definition and model learning, with more emphasis on reviewing, differentiating
and connecting different learning algorithms. We also discuss promising future
research directions.
|
Multi-Objective Cognitive Model: a supervised approach for multi-subject
fMRI analysis
|
In order to decode the human brain, Multivariate Pattern (MVP) classification
generates cognitive models by using functional Magnetic Resonance Imaging
(fMRI) datasets. As a standard pipeline in the MVP analysis, brain patterns in
multi-subject fMRI dataset must be mapped to a shared space and then a
classification model is generated by employing the mapped patterns. However,
the MVP models may not provide stable performance on a new fMRI dataset because
the standard pipeline uses disjoint steps for generating these models. Indeed,
each step in the pipeline includes an objective function with independent
optimization approach, where the best solution of each step may not be optimum
for the next steps. For tackling the mentioned issue, this paper introduces the
Multi-Objective Cognitive Model (MOCM) that utilizes an integrated objective
function for MVP analysis rather than just using those disjoint steps. For
solving the integrated problem, we proposed a customized multi-objective
optimization approach, where all possible solutions are firstly generated, and
then our method ranks and selects the robust solutions as the final results.
Empirical studies confirm that the proposed method can generate superior
performance in comparison with other techniques.
|
Combining Graph-based Dependency Features with Convolutional Neural
Network for Answer Triggering
|
Answer triggering is the task of selecting the best-suited answer for a given
question from a set of candidate answers if exists. In this paper, we present a
hybrid deep learning model for answer triggering, which combines several
dependency graph based alignment features, namely graph edit distance,
graph-based similarity and dependency graph coverage, with dense vector
embeddings from a Convolutional Neural Network (CNN). Our experiments on the
WikiQA dataset show that such a combination can more accurately trigger a
candidate answer compared to the previous state-of-the-art models. Comparative
study on WikiQA dataset shows 5.86% absolute F-score improvement at the
question level.
|
Structured Adversarial Attack: Towards General Implementation and Better
Interpretability
|
When generating adversarial examples to attack deep neural networks (DNNs),
Lp norm of the added perturbation is usually used to measure the similarity
between original image and adversarial example. However, such adversarial
attacks perturbing the raw input spaces may fail to capture structural
information hidden in the input. This work develops a more general attack
model, i.e., the structured attack (StrAttack), which explores group sparsity
in adversarial perturbations by sliding a mask through images aiming for
extracting key spatial structures. An ADMM (alternating direction method of
multipliers)-based framework is proposed that can split the original problem
into a sequence of analytically solvable subproblems and can be generalized to
implement other attacking methods. Strong group sparsity is achieved in
adversarial perturbations even with the same level of Lp norm distortion as the
state-of-the-art attacks. We demonstrate the effectiveness of StrAttack by
extensive experimental results onMNIST, CIFAR-10, and ImageNet. We also show
that StrAttack provides better interpretability (i.e., better correspondence
with discriminative image regions)through adversarial saliency map (Papernot et
al., 2016b) and class activation map(Zhou et al., 2016).
|
Missing Value Imputation Based on Deep Generative Models
|
Missing values widely exist in many real-world datasets, which hinders the
performing of advanced data analytics. Properly filling these missing values is
crucial but challenging, especially when the missing rate is high. Many
approaches have been proposed for missing value imputation (MVI), but they are
mostly heuristics-based, lacking a principled foundation and do not perform
satisfactorily in practice. In this paper, we propose a probabilistic framework
based on deep generative models for MVI. Under this framework, imputing the
missing entries amounts to seeking a fixed-point solution between two
conditional distributions defined on the missing entries and latent variables
respectively. These distributions are parameterized by deep neural networks
(DNNs) which possess high approximation power and can capture the nonlinear
relationships between missing entries and the observed values. The learning of
weight parameters of DNNs is performed by maximizing an approximation of the
log-likelihood of observed values. We conducted extensive evaluation on 13
datasets and compared with 11 baselines methods, where our methods largely
outperforms the baselines.
|
Too many secants: a hierarchical approach to secant-based dimensionality
reduction on large data sets
|
A fundamental question in many data analysis settings is the problem of
discerning the "natural" dimension of a data set. That is, when a data set is
drawn from a manifold (possibly with noise), a meaningful aspect of the data is
the dimension of that manifold. Various approaches exist for estimating this
dimension, such as the method of Secant-Avoidance Projection (SAP).
Intuitively, the SAP algorithm seeks to determine a projection which best
preserves the lengths of all secants between points in a data set; by applying
the algorithm to find the best projections to vector spaces of various
dimensions, one may infer the dimension of the manifold of origination. That
is, one may learn the dimension at which it is possible to construct a
diffeomorphic copy of the data in a lower-dimensional Euclidean space. Using
Whitney's embedding theorem, we can relate this information to the natural
dimension of the data. A drawback of the SAP algorithm is that a data set with
$T$ points has $O(T^2)$ secants, making the computation and storage of all
secants infeasible for very large data sets. In this paper, we propose a novel
algorithm that generalizes the SAP algorithm with an emphasis on addressing
this issue. That is, we propose a hierarchical secant-based
dimensionality-reduction method, which can be employed for data sets where
explicitly calculating all secants is not feasible.
|
Hybrid Subspace Learning for High-Dimensional Data
|
The high-dimensional data setting, in which p >> n, is a challenging
statistical paradigm that appears in many real-world problems. In this setting,
learning a compact, low-dimensional representation of the data can
substantially help distinguish signal from noise. One way to achieve this goal
is to perform subspace learning to estimate a small set of latent features that
capture the majority of the variance in the original data. Most existing
subspace learning models, such as PCA, assume that the data can be fully
represented by its embedding in one or more latent subspaces. However, in this
work, we argue that this assumption is not suitable for many high-dimensional
datasets; often only some variables can easily be projected to a
low-dimensional space. We propose a hybrid dimensionality reduction technique
in which some features are mapped to a low-dimensional subspace while others
remain in the original space. Our model leads to more accurate estimation of
the latent space and lower reconstruction error. We present a simple
optimization procedure for the resulting biconvex problem and show synthetic
data results that demonstrate the advantages of our approach over existing
methods. Finally, we demonstrate the effectiveness of this method for
extracting meaningful features from both gene expression and video background
subtraction datasets.
|
Concentration bounds for empirical conditional value-at-risk: The
unbounded case
|
In several real-world applications involving decision making under
uncertainty, the traditional expected value objective may not be suitable, as
it may be necessary to control losses in the case of a rare but extreme event.
Conditional Value-at-Risk (CVaR) is a popular risk measure for modeling the
aforementioned objective. We consider the problem of estimating CVaR from
i.i.d. samples of an unbounded random variable, which is either sub-Gaussian or
sub-exponential. We derive a novel one-sided concentration bound for a natural
sample-based CVaR estimator in this setting. Our bound relies on a
concentration result for a quantile-based estimator for Value-at-Risk (VaR),
which may be of independent interest.
|
NIMFA: A Python Library for Nonnegative Matrix Factorization
|
NIMFA is an open-source Python library that provides a unified interface to
nonnegative matrix factorization algorithms. It includes implementations of
state-of-the-art factorization methods, initialization approaches, and quality
scoring. It supports both dense and sparse matrix representation. NIMFA's
component-based implementation and hierarchical design should help the users to
employ already implemented techniques or design and code new strategies for
matrix factorization tasks.
|
Gray-box Adversarial Training
|
Adversarial samples are perturbed inputs crafted to mislead the machine
learning systems. A training mechanism, called adversarial training, which
presents adversarial samples along with clean samples has been introduced to
learn robust models. In order to scale adversarial training for large datasets,
these perturbations can only be crafted using fast and simple methods (e.g.,
gradient ascent). However, it is shown that adversarial training converges to a
degenerate minimum, where the model appears to be robust by generating weaker
adversaries. As a result, the models are vulnerable to simple black-box
attacks. In this paper we, (i) demonstrate the shortcomings of existing
evaluation policy, (ii) introduce novel variants of white-box and black-box
attacks, dubbed gray-box adversarial attacks" based on which we propose novel
evaluation method to assess the robustness of the learned models, and (iii)
propose a novel variant of adversarial training, named Graybox Adversarial
Training" that uses intermediate versions of the models to seed the
adversaries. Experimental evaluation demonstrates that the models trained using
our method exhibit better robustness compared to both undefended and
adversarially trained model
|
Regret Bounds for Reinforcement Learning via Markov Chain Concentration
|
We give a simple optimistic algorithm for which it is easy to derive regret
bounds of $\tilde{O}(\sqrt{t_{\rm mix} SAT})$ after $T$ steps in uniformly
ergodic Markov decision processes with $S$ states, $A$ actions, and mixing time
parameter $t_{\rm mix}$. These bounds are the first regret bounds in the
general, non-episodic setting with an optimal dependence on all given
parameters. They could only be improved by using an alternative mixing time
parameter.
|
Beyond $1/2$-Approximation for Submodular Maximization on Massive Data
Streams
|
Many tasks in machine learning and data mining, such as data diversification,
non-parametric learning, kernel machines, clustering etc., require extracting a
small but representative summary from a massive dataset. Often, such problems
can be posed as maximizing a submodular set function subject to a cardinality
constraint. We consider this question in the streaming setting, where elements
arrive over time at a fast pace and thus we need to design an efficient,
low-memory algorithm. One such method, proposed by Badanidiyuru et al. (2014),
always finds a $0.5$-approximate solution. Can this approximation factor be
improved? We answer this question affirmatively by designing a new algorithm
SALSA for streaming submodular maximization. It is the first low-memory,
single-pass algorithm that improves the factor $0.5$, under the natural
assumption that elements arrive in a random order. We also show that this
assumption is necessary, i.e., that there is no such algorithm with better than
$0.5$-approximation when elements arrive in arbitrary order. Our experiments
demonstrate that SALSA significantly outperforms the state of the art in
applications related to exemplar-based clustering, social graph analysis, and
recommender systems.
|
Statistical Windows in Testing for the Initial Distribution of a
Reversible Markov Chain
|
We study the problem of hypothesis testing between two discrete
distributions, where we only have access to samples after the action of a known
reversible Markov chain, playing the role of noise. We derive
instance-dependent minimax rates for the sample complexity of this problem, and
show how its dependence in time is related to the spectral properties of the
Markov chain. We show that there exists a wide statistical window, in terms of
sample complexity for hypothesis testing between different pairs of initial
distributions. We illustrate these results in several concrete examples.
|
Residual Memory Networks: Feed-forward approach to learn long temporal
dependencies
|
Training deep recurrent neural network (RNN) architectures is complicated due
to the increased network complexity. This disrupts the learning of higher order
abstracts using deep RNN. In case of feed-forward networks training deep
structures is simple and faster while learning long-term temporal information
is not possible. In this paper we propose a residual memory neural network
(RMN) architecture to model short-time dependencies using deep feed-forward
layers having residual and time delayed connections. The residual connection
paves way to construct deeper networks by enabling unhindered flow of gradients
and the time delay units capture temporal information with shared weights. The
number of layers in RMN signifies both the hierarchical processing depth and
temporal depth. The computational complexity in training RMN is significantly
less when compared to deep recurrent networks. RMN is further extended as
bi-directional RMN (BRMN) to capture both past and future information.
Experimental analysis is done on AMI corpus to substantiate the capability of
RMN in learning long-term information and hierarchical information. Recognition
performance of RMN trained with 300 hours of Switchboard corpus is compared
with various state-of-the-art LVCSR systems. The results indicate that RMN and
BRMN gains 6 % and 3.8 % relative improvement over LSTM and BLSTM networks.
|
V-FCNN: Volumetric Fully Convolution Neural Network For Automatic Atrial
Segmentation
|
Atrial Fibrillation (AF) is a common electro-physiological cardiac disorder
that causes changes in the anatomy of the atria. A better characterization of
these changes is desirable for the definition of clinical biomarkers,
furthermore, thus there is a need for its fully automatic segmentation from
clinical images. In this work, we present an architecture based on
3D-convolution kernels, a Volumetric Fully Convolution Neural Network (V-FCNN),
able to segment the entire volume in a one-shot, and consequently integrate the
implicit spatial redundancy present in high-resolution images. A loss function
based on the mixture of both Mean Square Error (MSE) and Dice Loss (DL) is
used, in an attempt to combine the ability to capture the bulk shape as well as
the reduction of local errors products by over-segmentation. Results
demonstrate a reasonable performance in the middle region of the atria along
with the impact of the challenges of capturing the variability of the pulmonary
veins or the identification of the valve plane that separates the atria to the
ventricle. A final dice of $92.5\%$ in $54$ patients ($4752$ atria test slices
in total) is shown.
|
Distributional Multivariate Policy Evaluation and Exploration with the
Bellman GAN
|
The recently proposed distributional approach to reinforcement learning
(DiRL) is centered on learning the distribution of the reward-to-go, often
referred to as the value distribution. In this work, we show that the
distributional Bellman equation, which drives DiRL methods, is equivalent to a
generative adversarial network (GAN) model. In this formulation, DiRL can be
seen as learning a deep generative model of the value distribution, driven by
the discrepancy between the distribution of the current value, and the
distribution of the sum of current reward and next value. We use this insight
to propose a GAN-based approach to DiRL, which leverages the strengths of GANs
in learning distributions of high-dimensional data. In particular, we show that
our GAN approach can be used for DiRL with multivariate rewards, an important
setting which cannot be tackled with prior methods. The multivariate setting
also allows us to unify learning the distribution of values and state
transitions, and we exploit this idea to devise a novel exploration method that
is driven by the discrepancy in estimating both values and states.
|
A Survey on Deep Transfer Learning
|
As a new classification platform, deep learning has recently received
increasing attention from researchers and has been successfully applied to many
domains. In some domains, like bioinformatics and robotics, it is very
difficult to construct a large-scale well-annotated dataset due to the expense
of data acquisition and costly annotation, which limits its development.
Transfer learning relaxes the hypothesis that the training data must be
independent and identically distributed (i.i.d.) with the test data, which
motivates us to use transfer learning to solve the problem of insufficient
training data. This survey focuses on reviewing the current researches of
transfer learning by using deep neural network and its applications. We defined
deep transfer learning, category and review the recent research works based on
the techniques used in deep transfer learning.
|
A Survey on Surrogate Approaches to Non-negative Matrix Factorization
|
Motivated by applications in hyperspectral imaging we investigate methods for
approximating a high-dimensional non-negative matrix $\mathbf{\mathit{Y}}$ by a
product of two lower-dimensional, non-negative matrices $\mathbf{\mathit{K}}$
and $\mathbf{\mathit{X}}.$ This so-called non-negative matrix factorization is
based on defining suitable Tikhonov functionals, which combine a discrepancy
measure for $\mathbf{\mathit{Y}}\approx\mathbf{\mathit{KX}}$ with penalty terms
for enforcing additional properties of $\mathbf{\mathit{K}}$ and
$\mathbf{\mathit{X}}$. The minimization is based on alternating minimization
with respect to $\mathbf{\mathit{K}}$ or $\mathbf{\mathit{X}}$, where in each
iteration step one replaces the original Tikhonov functional by a locally
defined surrogate functional. The choice of surrogate functionals is crucial:
It should allow a comparatively simple minimization and simultaneously its
first order optimality condition should lead to multiplicative update rules,
which automatically preserve non-negativity of the iterates. We review the most
standard construction principles for surrogate functionals for Frobenius-norm
and Kullback-Leibler discrepancy measures. We extend the known surrogate
constructions by a general framework, which allows to add a large variety of
penalty terms. The paper finishes by deriving the corresponding alternating
minimization schemes explicitely and by applying these methods to MALDI imaging
data.
|
Adversarial Vision Challenge
|
The NIPS 2018 Adversarial Vision Challenge is a competition to facilitate
measurable progress towards robust machine vision models and more generally
applicable adversarial attacks. This document is an updated version of our
competition proposal that was accepted in the competition track of 32nd
Conference on Neural Information Processing Systems (NIPS 2018).
|
Hashing with Binary Matrix Pursuit
|
We propose theoretical and empirical improvements for two-stage hashing
methods. We first provide a theoretical analysis on the quality of the binary
codes and show that, under mild assumptions, a residual learning scheme can
construct binary codes that fit any neighborhood structure with arbitrary
accuracy. Secondly, we show that with high-capacity hash functions such as
CNNs, binary code inference can be greatly simplified for many standard
neighborhood definitions, yielding smaller optimization problems and more
robust codes. Incorporating our findings, we propose a novel two-stage hashing
method that significantly outperforms previous hashing studies on widely used
image retrieval benchmarks.
|
Simultaneous Edge Alignment and Learning
|
Edge detection is among the most fundamental vision problems for its role in
perceptual grouping and its wide applications. Recent advances in
representation learning have led to considerable improvements in this area.
Many state of the art edge detection models are learned with fully
convolutional networks (FCNs). However, FCN-based edge learning tends to be
vulnerable to misaligned labels due to the delicate structure of edges. While
such problem was considered in evaluation benchmarks, similar issue has not
been explicitly addressed in general edge learning. In this paper, we show that
label misalignment can cause considerably degraded edge learning quality, and
address this issue by proposing a simultaneous edge alignment and learning
framework. To this end, we formulate a probabilistic model where edge alignment
is treated as latent variable optimization, and is learned end-to-end during
network training. Experiments show several applications of this work, including
improved edge detection with state of the art performance, and automatic
refinement of noisy annotations.
|
MCRM: Mother Compact Recurrent Memory
|
LSTMs and GRUs are the most common recurrent neural network architectures
used to solve temporal sequence problems. The two architectures have differing
data flows dealing with a common component called the cell state (also referred
to as the memory). We attempt to enhance the memory by presenting a
modification that we call the Mother Compact Recurrent Memory (MCRM). MCRMs are
a type of a nested LSTM-GRU architecture where the cell state is the GRU hidden
state. The concatenation of the forget gate and input gate interactions from
the LSTM are considered an input to the GRU cell. Because MCRMs has this type
of nesting, MCRMs have a compact memory pattern consisting of neurons that acts
explicitly in both long-term and short-term fashions. For some specific tasks,
empirical results show that MCRMs outperform previously used architectures.
|
Withholding or withdrawing invasive interventions may not accelerate
time to death among dying ICU patients
|
We considered observational data available from the MIMIC-III open-access ICU
database and collected within a study period between year 2002 up to 2011. If a
patient had multiple admissions to the ICU during the 30 days before death,
only the first stay was analyzed, leading to a final set of 6,436 unique ICU
admissions during the study period. We tested two hypotheses: (i)
administration of invasive intervention during the ICU stay immediately
preceding end-of-life would decrease over the study time period and (ii)
time-to-death from ICU admission would also decrease, due to the decrease in
invasive intervention administration. To investigate the latter hypothesis, we
performed a subgroups analysis by considering patients with lowest and highest
severity. To do so, we stratified the patients based on their SAPS I scores,
and we considered patients within the first and the third tertiles of the
score. We then assessed differences in trends within these groups between years
2002-05 vs. 2008-11.
Comparing the period 2002-2005 vs. 2008-2011, we found a reduction in
endotracheal ventilation among patients who died within 30 days of ICU
admission (120.8 vs. 68.5 hours for the lowest severity patients, p<0.001; 47.7
vs. 46.0 hours for the highest severity patients, p=0.004). This is explained
in part by an increase in the use of non-invasive ventilation. Comparing the
period 2002-2005 vs. 2008-2011, we found a reduction in the use of vasopressors
and inotropes among patients with the lowest severity who died within 30 days
of ICU admission (41.8 vs. 36.2 hours, p<0.001) but not among those with the
highest severity. Despite a reduction in the use of invasive interventions, we
did not find a reduction in the time to death between 2002-2005 vs. 2008-2011
(7.8 days vs. 8.2 days for the lowest severity patients, p=0.32; 2.1 days vs.
2.0 days for the highest severity patients, p=0.74).
|
Outlier detection on network flow analysis
|
It is important to be able to detect and classify malicious network traffic
flows such as DDoS attacks from benign flows. Normally the task is performed by
using supervised classification algorithms. In this paper we analyze the usage
of outlier detection algorithms for the network traffic classification problem.
|
Active Learning based on Data Uncertainty and Model Sensitivity
|
Robots can rapidly acquire new skills from demonstrations. However, during
generalisation of skills or transitioning across fundamentally different
skills, it is unclear whether the robot has the necessary knowledge to perform
the task. Failing to detect missing information often leads to abrupt movements
or to collisions with the environment. Active learning can quantify the
uncertainty of performing the task and, in general, locate regions of missing
information. We introduce a novel algorithm for active learning and demonstrate
its utility for generating smooth trajectories. Our approach is based on deep
generative models and metric learning in latent spaces. It relies on the
Jacobian of the likelihood to detect non-smooth transitions in the latent
space, i.e., transitions that lead to abrupt changes in the movement of the
robot. When non-smooth transitions are detected, our algorithm asks for an
additional demonstration from that specific region. The newly acquired
knowledge modifies the data manifold and allows for learning a latent
representation for generating smooth movements. We demonstrate the efficacy of
our approach on generalising elementary skills, transitioning across different
skills, and implicitly avoiding collisions with the environment. For our
experiments, we use a simulated pendulum where we observe its motion from
images and a 7-DoF anthropomorphic arm.
|
Unbiased Implicit Variational Inference
|
We develop unbiased implicit variational inference (UIVI), a method that
expands the applicability of variational inference by defining an expressive
variational family. UIVI considers an implicit variational distribution
obtained in a hierarchical manner using a simple reparameterizable distribution
whose variational parameters are defined by arbitrarily flexible deep neural
networks. Unlike previous works, UIVI directly optimizes the evidence lower
bound (ELBO) rather than an approximation to the ELBO. We demonstrate UIVI on
several models, including Bayesian multinomial logistic regression and
variational autoencoders, and show that UIVI achieves both tighter ELBO and
better predictive performance than existing approaches at a similar
computational cost.
|
Efficient Principal Subspace Projection of Streaming Data Through Fast
Similarity Matching
|
Big data problems frequently require processing datasets in a streaming
fashion, either because all data are available at once but collectively are
larger than available memory or because the data intrinsically arrive one data
point at a time and must be processed online. Here, we introduce a
computationally efficient version of similarity matching, a framework for
online dimensionality reduction that incrementally estimates the top
K-dimensional principal subspace of streamed data while keeping in memory only
the last sample and the current iterate. To assess the performance of our
approach, we construct and make public a test suite containing both a synthetic
data generator and the infrastructure to test online dimensionality reduction
algorithms on real datasets, as well as performant implementations of our
algorithm and competing algorithms with similar aims. Among the algorithms
considered we find our approach to be competitive, performing among the best on
both synthetic and real data.
|
Learning to Share and Hide Intentions using Information Regularization
|
Learning to cooperate with friends and compete with foes is a key component
of multi-agent reinforcement learning. Typically to do so, one requires access
to either a model of or interaction with the other agent(s). Here we show how
to learn effective strategies for cooperation and competition in an asymmetric
information game with no such model or interaction. Our approach is to
encourage an agent to reveal or hide their intentions using an
information-theoretic regularizer. We consider both the mutual information
between goal and action given state, as well as the mutual information between
goal and state. We show how to optimize these regularizers in a way that is
easy to integrate with policy gradient reinforcement learning. Finally, we
demonstrate that cooperative (competitive) policies learned with our approach
lead to more (less) reward for a second agent in two simple asymmetric
information games.
|
Semi-supervised Deep Generative Modelling of Incomplete Multi-Modality
Emotional Data
|
There are threefold challenges in emotion recognition. First, it is difficult
to recognize human's emotional states only considering a single modality.
Second, it is expensive to manually annotate the emotional data. Third,
emotional data often suffers from missing modalities due to unforeseeable
sensor malfunction or configuration issues. In this paper, we address all these
problems under a novel multi-view deep generative framework. Specifically, we
propose to model the statistical relationships of multi-modality emotional data
using multiple modality-specific generative networks with a shared latent
space. By imposing a Gaussian mixture assumption on the posterior approximation
of the shared latent variables, our framework can learn the joint deep
representation from multiple modalities and evaluate the importance of each
modality simultaneously. To solve the labeled-data-scarcity problem, we extend
our multi-view model to semi-supervised learning scenario by casting the
semi-supervised classification problem as a specialized missing data imputation
task. To address the missing-modality problem, we further extend our
semi-supervised multi-view model to deal with incomplete data, where a missing
view is treated as a latent variable and integrated out during inference. This
way, the proposed overall framework can utilize all available (both labeled and
unlabeled, as well as both complete and incomplete) data to improve its
generalization ability. The experiments conducted on two real multi-modal
emotion datasets demonstrated the superiority of our framework.
|
Paying Attention to Attention: Highlighting Influential Samples in
Sequential Analysis
|
In (Yang et al. 2016), a hierarchical attention network (HAN) is created for
document classification. The attention layer can be used to visualize text
influential in classifying the document, thereby explaining the model's
prediction. We successfully applied HAN to a sequential analysis task in the
form of real-time monitoring of turn taking in conversations. However, we
discovered instances where the attention weights were uniform at the stopping
point (indicating all turns were equivalently influential to the classifier),
preventing meaningful visualization for real-time human review or classifier
improvement. We observed that attention weights for turns fluctuated as the
conversations progressed, indicating turns had varying influence based on
conversation state. Leveraging this observation, we develop a method to create
more informative real-time visuals (as confirmed by human reviewers) in cases
of uniform attention weights using the changes in turn importance as a
conversation progresses over time.
|
Structure Learning for Relational Logistic Regression: An Ensemble
Approach
|
We consider the problem of learning Relational Logistic Regression (RLR).
Unlike standard logistic regression, the features of RLRs are first-order
formulae with associated weight vectors instead of scalar weights. We turn the
problem of learning RLR to learning these vector-weighted formulae and develop
a learning algorithm based on the recently successful functional-gradient
boosting methods for probabilistic logic models. We derive the functional
gradients and show how weights can be learned simultaneously in an efficient
manner. Our empirical evaluation on standard and novel data sets demonstrates
the superiority of our approach over other methods for learning RLR.
|
Probabilistic Causal Analysis of Social Influence
|
Mastering the dynamics of social influence requires separating, in a database
of information propagation traces, the genuine causal processes from temporal
correlation, i.e., homophily and other spurious causes. However, most studies
to characterize social influence, and, in general, most data-science analyses
focus on correlations, statistical independence, or conditional independence.
Only recently, there has been a resurgence of interest in "causal data
science", e.g., grounded on causality theories. In this paper we adopt a
principled causal approach to the analysis of social influence from
information-propagation data, rooted in the theory of probabilistic causation.
Our approach consists of two phases. In the first one, in order to avoid the
pitfalls of misinterpreting causation when the data spans a mixture of several
subtypes ("Simpson's paradox"), we partition the set of propagation traces into
groups, in such a way that each group is as less contradictory as possible in
terms of the hierarchical structure of information propagation. To achieve this
goal, we borrow the notion of "agony" and define the Agony-bounded Partitioning
problem, which we prove being hard, and for which we develop two efficient
algorithms with approximation guarantees. In the second phase, for each group
from the first phase, we apply a constrained MLE approach to ultimately learn a
minimal causal topology. Experiments on synthetic data show that our method is
able to retrieve the genuine causal arcs w.r.t. a ground-truth generative
model. Experiments on real data show that, by focusing only on the extracted
causal structures instead of the whole social graph, the effectiveness of
predicting influence spread is significantly improved.
|
Fast Variance Reduction Method with Stochastic Batch Size
|
In this paper we study a family of variance reduction methods with randomized
batch size---at each step, the algorithm first randomly chooses the batch size
and then selects a batch of samples to conduct a variance-reduced stochastic
update. We give the linear convergence rate for this framework for composite
functions, and show that the optimal strategy to achieve the optimal
convergence rate per data access is to always choose batch size of 1, which is
equivalent to the SAGA algorithm. However, due to the presence of cache/disk IO
effect in computer architecture, the number of data access cannot reflect the
running time because of 1) random memory access is much slower than sequential
access, 2) when data is too big to fit into memory, disk seeking takes even
longer time. After taking these into account, choosing batch size of $1$ is no
longer optimal, so we propose a new algorithm called SAGA++ and show how to
calculate the optimal average batch size theoretically. Our algorithm
outperforms SAGA and other existing batched and stochastic solvers on real
datasets. In addition, we also conduct a precise analysis to compare different
update rules for variance reduction methods, showing that SAGA++ converges
faster than SVRG in theory.
|
Test without Trust: Optimal Locally Private Distribution Testing
|
We study the problem of distribution testing when the samples can only be
accessed using a locally differentially private mechanism and focus on two
representative testing questions of identity (goodness-of-fit) and independence
testing for discrete distributions. We are concerned with two settings: First,
when we insist on using an already deployed, general-purpose locally
differentially private mechanism such as the popular RAPPOR or the recently
introduced Hadamard Response for collecting data, and must build our tests
based on the data collected via this mechanism; and second, when no such
restriction is imposed, and we can design a bespoke mechanism specifically for
testing. For the latter purpose, we introduce the Randomized Aggregated Private
Testing Optimal Response (RAPTOR) mechanism which is remarkably simple and
requires only one bit of communication per sample.
We propose tests based on these mechanisms and analyze their sample
complexities. Each proposed test can be implemented efficiently. In each case
(barring one), we complement our performance bounds for algorithms with
information-theoretic lower bounds and establish sample optimality of our
proposed algorithm. A peculiar feature that emerges is that our sample-optimal
algorithm based on RAPTOR uses public-coins, and any test based on RAPPOR or
Hadamard Response, which are both private-coin mechanisms, requires
significantly more samples.
|
Instance-Dependent PU Learning by Bayesian Optimal Relabeling
|
When learning from positive and unlabelled data, it is a strong assumption
that the positive observations are randomly sampled from the distribution of
$X$ conditional on $Y = 1$, where X stands for the feature and Y the label.
Most existing algorithms are optimally designed under the assumption. However,
for many real-world applications, the observed positive examples are dependent
on the conditional probability $P(Y = 1|X)$ and should be sampled biasedly. In
this paper, we assume that a positive example with a higher $P(Y = 1|X)$ is
more likely to be labelled and propose a probabilistic-gap based PU learning
algorithms. Specifically, by treating the unlabelled data as noisy negative
examples, we could automatically label a group positive and negative examples
whose labels are identical to the ones assigned by a Bayesian optimal
classifier with a consistency guarantee. The relabelled examples have a biased
domain, which is remedied by the kernel mean matching technique. The proposed
algorithm is model-free and thus do not have any parameters to tune.
Experimental results demonstrate that our method works well on both generated
and real-world datasets.
|
Importance of the Mathematical Foundations of Machine Learning Methods
for Scientific and Engineering Applications
|
There has been a lot of recent interest in adopting machine learning methods
for scientific and engineering applications. This has in large part been
inspired by recent successes and advances in the domains of Natural Language
Processing (NLP) and Image Classification (IC). However, scientific and
engineering problems have their own unique characteristics and requirements
raising new challenges for effective design and deployment of machine learning
approaches. There is a strong need for further mathematical developments on the
foundations of machine learning methods to increase the level of rigor of
employed methods and to ensure more reliable and interpretable results. Also as
reported in the recent literature on state-of-the-art results and indicated by
the No Free Lunch Theorems of statistical learning theory incorporating some
form of inductive bias and domain knowledge is essential to success.
Consequently, even for existing and widely used methods there is a strong need
for further mathematical work to facilitate ways to incorporate prior
scientific knowledge and related inductive biases into learning frameworks and
algorithms. We briefly discuss these topics and discuss some ideas proceeding
in this direction.
|
Hierarchical Clustering better than Average-Linkage
|
Hierarchical Clustering (HC) is a widely studied problem in exploratory data
analysis, usually tackled by simple agglomerative procedures like
average-linkage, single-linkage or complete-linkage. In this paper we focus on
two objectives, introduced recently to give insight into the performance of
average-linkage clustering: a similarity based HC objective proposed by
[Moseley and Wang, 2017] and a dissimilarity based HC objective proposed by
[Cohen-Addad et al., 2018]. In both cases, we present tight counterexamples
showing that average-linkage cannot obtain better than 1/3 and 2/3
approximations respectively (in the worst-case), settling an open question
raised in [Moseley and Wang, 2017]. This matches the approximation ratio of a
random solution, raising a natural question: can we beat average-linkage for
these objectives? We answer this in the affirmative, giving two new algorithms
based on semidefinite programming with provably better guarantees.
|
Grassmannian Learning: Embedding Geometry Awareness in Shallow and Deep
Learning
|
Modern machine learning algorithms have been adopted in a range of
signal-processing applications spanning computer vision, natural language
processing, and artificial intelligence. Many relevant problems involve
subspace-structured features, orthogonality constrained or low-rank constrained
objective functions, or subspace distances. These mathematical characteristics
are expressed naturally using the Grassmann manifold. Unfortunately, this fact
is not yet explored in many traditional learning algorithms. In the last few
years, there have been growing interests in studying Grassmann manifold to
tackle new learning problems. Such attempts have been reassured by substantial
performance improvements in both classic learning and learning using deep
neural networks. We term the former as shallow and the latter deep Grassmannian
learning. The aim of this paper is to introduce the emerging area of
Grassmannian learning by surveying common mathematical problems and primary
solution approaches, and overviewing various applications. We hope to inspire
practitioners in different fields to adopt the powerful tool of Grassmannian
learning in their research.
|
Deep Stacked Stochastic Configuration Networks for Lifelong Learning of
Non-Stationary Data Streams
|
The concept of SCN offers a fast framework with universal approximation
guarantee for lifelong learning of non-stationary data streams. Its adaptive
scope selection property enables for proper random generation of hidden unit
parameters advancing conventional randomized approaches constrained with a
fixed scope of random parameters. This paper proposes deep stacked stochastic
configuration network (DSSCN) for continual learning of non-stationary data
streams which contributes two major aspects: 1) DSSCN features a
self-constructing methodology of deep stacked network structure where hidden
unit and hidden layer are extracted automatically from continuously generated
data streams; 2) the concept of SCN is developed to randomly assign inverse
covariance matrix of multivariate Gaussian function in the hidden node addition
step bypassing its computationally prohibitive tuning phase. Numerical
evaluation and comparison with prominent data stream algorithms under two
procedures: periodic hold-out and prequential test-then-train processes
demonstrate the advantage of proposed methodology.
|
DeePathology: Deep Multi-Task Learning for Inferring Molecular Pathology
from Cancer Transcriptome
|
Despite great advances, molecular cancer pathology is often limited to the
use of a small number of biomarkers rather than the whole transcriptome, partly
due to computational challenges. Here, we introduce a novel architecture of
Deep Neural Networks (DNNs) that is capable of simultaneous inference of
various properties of biological samples, through multi-task and transfer
learning. It encodes the whole transcription profile into a strikingly
low-dimensional latent vector of size 8, and then recovers mRNA and miRNA
expression profiles, tissue and disease type from this vector. This latent
space is significantly better than the original gene expression profiles for
discriminating samples based on their tissue and disease. We employed this
architecture on mRNA transcription profiles of 10787 clinical samples from 34
classes (one healthy and 33 different types of cancer) from 27 tissues. Our
method significantly outperforms prior works and classical machine learning
approaches in predicting tissue-of-origin, normal or disease state and cancer
type of each sample. For tissues with more than one type of cancer, it reaches
99.4\% accuracy in identifying the correct cancer subtype. We also show this
system is very robust against noise and missing values. Collectively, our
results highlight applications of artificial intelligence in molecular cancer
pathology and oncological research. DeePathology is freely available at
\url{https://github.com/SharifBioinf/DeePathology}.
|
Speeding Up Distributed Gradient Descent by Utilizing Non-persistent
Stragglers
|
Distributed gradient descent (DGD) is an efficient way of implementing
gradient descent (GD), especially for large data sets, by dividing the
computation tasks into smaller subtasks and assigning to different computing
servers (CSs) to be executed in parallel. In standard parallel execution,
per-iteration waiting time is limited by the execution time of the straggling
servers. Coded DGD techniques have been introduced recently, which can tolerate
straggling servers via assigning redundant computation tasks to the CSs. In
most of the existing DGD schemes, either with coded computation or coded
communication, the non-straggling CSs transmit one message per iteration once
they complete all their assigned computation tasks. However, although the
straggling servers cannot complete all their assigned tasks, they are often
able to complete a certain portion of them. In this paper, we allow multiple
transmissions from each CS at each iteration in order to make sure a maximum
number of completed computations can be reported to the aggregating server
(AS), including the straggling servers. We numerically show that the average
completion time per iteration can be reduced significantly by slightly
increasing the communication load per server.
|
Multi-Output Convolution Spectral Mixture for Gaussian Processes
|
Multi-output Gaussian processes (MOGPs) are an extension of Gaussian
Processes (GPs) for predicting multiple output variables (also called channels,
tasks) simultaneously. In this paper we use the convolution theorem to design a
new kernel for MOGPs, by modeling cross channel dependencies through cross
convolution of time and phase delayed components in the spectral domain. The
resulting kernel is called Multi-Output Convolution Spectral Mixture (MOCSM)
kernel. Results of extensive experiments on synthetic and real-life datasets
demonstrate the advantages of the proposed kernel and its state of the art
performance. MOCSM enjoys the desirable property to reduce to the well known
Spectral Mixture (SM) kernel when a single-channel is considered. A comparison
with the recently introduced Multi-Output Spectral Mixture kernel reveals that
this is not the case for the latter kernel, which contains quadratic terms that
generate undesirable scale effects when the spectral densities of different
channels are either very close or very far from each other in the frequency
domain.
|
A novel topology design approach using an integrated deep learning
network architecture
|
Topology design optimization offers tremendous opportunity in design and
manufacturing freedoms by designing and producing a part from the ground-up
without a meaningful initial design as required by conventional shape design
optimization approaches. Ideally, with adequate problem statements, to
formulate and solve the topology design problem using a standard topology
optimization process, such as SIMP (Simplified Isotropic Material with
Penalization) is possible. In reality, an estimated over thousands of design
iterations is often required for just a few design variables, the conventional
optimization approach is in general impractical or computationally unachievable
for real world applications significantly diluting the development of the
topology optimization technology. There is, therefore, a need for a different
approach that will be able to optimize the initial design topology effectively
and rapidly. Therefore, this work presents a new topology design procedure to
generate optimal structures using an integrated Generative Adversarial Networks
(GANs) and convolutional neural network architecture.
|
Acoustic Scene Classification: A Competition Review
|
In this paper we study the problem of acoustic scene classification, i.e.,
categorization of audio sequences into mutually exclusive classes based on
their spectral content. We describe the methods and results discovered during a
competition organized in the context of a graduate machine learning course;
both by the students and external participants. We identify the most suitable
methods and study the impact of each by performing an ablation study of the
mixture of approaches. We also compare the results with a neural network
baseline, and show the improvement over that. Finally, we discuss the impact of
using a competition as a part of a university course, and justify its
importance in the curriculum based on student feedback.
|
Application of End-to-End Deep Learning in Wireless Communications
Systems
|
Deep learning is a potential paradigm changer for the design of wireless
communications systems (WCS), from conventional handcrafted schemes based on
sophisticated mathematical models with assumptions to autonomous schemes based
on the end-to-end deep learning using a large number of data. In this article,
we present a basic concept of the deep learning and its application to WCS by
investigating the resource allocation (RA) scheme based on a deep neural
network (DNN) where multiple goals with various constraints can be satisfied
through the end-to-end deep learning. Especially, the optimality and
feasibility of the DNN based RA are verified through simulation. Then, we
discuss the technical challenges regarding the application of deep learning in
WCS.
|
Robust Implicit Backpropagation
|
Arguably the biggest challenge in applying neural networks is tuning the
hyperparameters, in particular the learning rate. The sensitivity to the
learning rate is due to the reliance on backpropagation to train the network.
In this paper we present the first application of Implicit Stochastic Gradient
Descent (ISGD) to train neural networks, a method known in convex optimization
to be unconditionally stable and robust to the learning rate. Our key
contribution is a novel layer-wise approximation of ISGD which makes its
updates tractable for neural networks. Experiments show that our method is more
robust to high learning rates and generally outperforms standard
backpropagation on a variety of tasks.
|
Mixed Integer Linear Programming for Feature Selection in Support Vector
Machine
|
This work focuses on support vector machine (SVM) with feature selection. A
MILP formulation is proposed for the problem. The choice of suitable features
to construct the separating hyperplanes has been modelled in this formulation
by including a budget constraint that sets in advance a limit on the number of
features to be used in the classification process. We propose both an exact and
a heuristic procedure to solve this formulation in an efficient way. Finally,
the validation of the model is done by checking it with some well-known data
sets and comparing it with classical classification methods.
|
The Sample Complexity of Up-to-$\varepsilon$ Multi-Dimensional Revenue
Maximization
|
We consider the sample complexity of revenue maximization for multiple
bidders in unrestricted multi-dimensional settings. Specifically, we study the
standard model of $n$ additive bidders whose values for $m$ heterogeneous items
are drawn independently. For any such instance and any $\varepsilon>0$, we show
that it is possible to learn an $\varepsilon$-Bayesian Incentive Compatible
auction whose expected revenue is within $\varepsilon$ of the optimal
$\varepsilon$-BIC auction from only polynomially many samples.
Our fully nonparametric approach is based on ideas that hold quite generally,
and completely sidestep the difficulty of characterizing optimal (or
near-optimal) auctions for these settings. Therefore, our results easily extend
to general multi-dimensional settings, including valuations that are not
necessarily even subadditive, and arbitrary allocation constraints. For the
cases of a single bidder and many goods, or a single parameter (good) and many
bidders, our analysis yields exact incentive compatibility (and for the latter
also computational efficiency). Although the single-parameter case is already
well-understood, our corollary for this case extends slightly the
state-of-the-art.
|
Multi-Label Zero-Shot Learning with Transfer-Aware Label Embedding
Projection
|
Zero-shot learning transfers knowledge from seen classes to novel unseen
classes to reduce human labor of labelling data for building new classifiers.
Much effort on zero-shot learning however has focused on the standard
multi-class setting, the more challenging multi-label zero-shot problem has
received limited attention. In this paper we propose a transfer-aware embedding
projection approach to tackle multi-label zero-shot learning. The approach
projects the label embedding vectors into a low-dimensional space to induce
better inter-label relationships and explicitly facilitate information transfer
from seen labels to unseen labels, while simultaneously learning a max-margin
multi-label classifier with the projected label embeddings. Auxiliary
information can be conveniently incorporated to guide the label embedding
projection to further improve label relation structures for zero-shot knowledge
transfer. We conduct experiments for zero-shot multi-label image
classification. The results demonstrate the efficacy of the proposed approach.
|
Deep context: end-to-end contextual speech recognition
|
In automatic speech recognition (ASR) what a user says depends on the
particular context she is in. Typically, this context is represented as a set
of word n-grams. In this work, we present a novel, all-neural, end-to-end (E2E)
ASR sys- tem that utilizes such context. Our approach, which we re- fer to as
Contextual Listen, Attend and Spell (CLAS) jointly- optimizes the ASR
components along with embeddings of the context n-grams. During inference, the
CLAS system can be presented with context phrases which might contain out-of-
vocabulary (OOV) terms not seen during training. We com- pare our proposed
system to a more traditional contextualiza- tion approach, which performs
shallow-fusion between inde- pendently trained LAS and contextual n-gram models
during beam search. Across a number of tasks, we find that the pro- posed CLAS
system outperforms the baseline method by as much as 68% relative WER,
indicating the advantage of joint optimization over individually trained
components. Index Terms: speech recognition, sequence-to-sequence models,
listen attend and spell, LAS, attention, embedded speech recognition.
|
Message Passing Graph Kernels
|
Graph kernels have recently emerged as a promising approach for tackling the
graph similarity and learning tasks at the same time. In this paper, we propose
a general framework for designing graph kernels. The proposed framework
capitalizes on the well-known message passing scheme on graphs. The kernels
derived from the framework consist of two components. The first component is a
kernel between vertices, while the second component is a kernel between graphs.
The main idea behind the proposed framework is that the representations of the
vertices are implicitly updated using an iterative procedure. Then, these
representations serve as the building blocks of a kernel that compares pairs of
graphs. We derive four instances of the proposed framework, and show through
extensive experiments that these instances are competitive with
state-of-the-art methods in various tasks.
|
Rethinking Numerical Representations for Deep Neural Networks
|
With ever-increasing computational demand for deep learning, it is critical
to investigate the implications of the numeric representation and precision of
DNN model weights and activations on computational efficiency. In this work, we
explore unconventional narrow-precision floating-point representations as it
relates to inference accuracy and efficiency to steer the improved design of
future DNN platforms. We show that inference using these custom numeric
representations on production-grade DNNs, including GoogLeNet and VGG, achieves
an average speedup of 7.6x with less than 1% degradation in inference accuracy
relative to a state-of-the-art baseline platform representing the most
sophisticated hardware using single-precision floating point. To facilitate the
use of such customized precision, we also present a novel technique that
drastically reduces the time required to derive the optimal precision
configuration.
|
Parallel and Streaming Algorithms for K-Core Decomposition
|
The $k$-core decomposition is a fundamental primitive in many machine
learning and data mining applications. We present the first distributed and the
first streaming algorithms to compute and maintain an approximate $k$-core
decomposition with provable guarantees. Our algorithms achieve rigorous bounds
on space complexity while bounding the number of passes or number of rounds of
computation. We do so by presenting a new powerful sketching technique for
$k$-core decomposition, and then by showing it can be computed efficiently in
both streaming and MapReduce models. Finally, we confirm the effectiveness of
our sketching technique empirically on a number of publicly available graphs.
|
PIVETed-Granite: Computational Phenotypes through Constrained Tensor
Factorization
|
It has been recently shown that sparse, nonnegative tensor factorization of
multi-modal electronic health record data is a promising approach to
high-throughput computational phenotyping. However, such approaches typically
do not leverage available domain knowledge while extracting the phenotypes;
hence, some of the suggested phenotypes may not map well to clinical concepts
or may be very similar to other suggested phenotypes. To address these issues,
we present a novel, automatic approach called PIVETed-Granite that mines
existing biomedical literature (PubMed) to obtain cannot-link constraints that
are then used as side-information during a tensor-factorization based
computational phenotyping process. The resulting improvements are clearly
observed in experiments using a large dataset from VUMC to identify phenotypes
for hypertensive patients.
|
L-Shapley and C-Shapley: Efficient Model Interpretation for Structured
Data
|
We study instancewise feature importance scoring as a method for model
interpretation. Any such method yields, for each predicted instance, a vector
of importance scores associated with the feature vector. Methods based on the
Shapley score have been proposed as a fair way of computing feature
attributions of this kind, but incur an exponential complexity in the number of
features. This combinatorial explosion arises from the definition of the
Shapley value and prevents these methods from being scalable to large data sets
and complex models. We focus on settings in which the data have a graph
structure, and the contribution of features to the target variable is
well-approximated by a graph-structured factorization. In such settings, we
develop two algorithms with linear complexity for instancewise feature
importance scoring. We establish the relationship of our methods to the Shapley
value and another closely related concept known as the Myerson value from
cooperative game theory. We demonstrate on both language and image data that
our algorithms compare favorably with other methods for model interpretation.
|
Courteous Autonomous Cars
|
Typically, autonomous cars optimize for a combination of safety, efficiency,
and driving quality. But as we get better at this optimization, we start seeing
behavior go from too conservative to too aggressive. The car's behavior exposes
the incentives we provide in its cost function. In this work, we argue for cars
that are not optimizing a purely selfish cost, but also try to be courteous to
other interactive drivers. We formalize courtesy as a term in the objective
that measures the increase in another driver's cost induced by the autonomous
car's behavior. Such a courtesy term enables the robot car to be aware of
possible irrationality of the human behavior, and plan accordingly. We analyze
the effect of courtesy in a variety of scenarios. We find, for example, that
courteous robot cars leave more space when merging in front of a human driver.
Moreover, we find that such a courtesy term can help explain real human driver
behavior on the NGSIM dataset.
|
Beyond Pixel Norm-Balls: Parametric Adversaries using an Analytically
Differentiable Renderer
|
Many machine learning image classifiers are vulnerable to adversarial
attacks, inputs with perturbations designed to intentionally trigger
misclassification. Current adversarial methods directly alter pixel colors and
evaluate against pixel norm-balls: pixel perturbations smaller than a specified
magnitude, according to a measurement norm. This evaluation, however, has
limited practical utility since perturbations in the pixel space do not
correspond to underlying real-world phenomena of image formation that lead to
them and has no security motivation attached. Pixels in natural images are
measurements of light that has interacted with the geometry of a physical
scene. As such, we propose the direct perturbation of physical parameters that
underly image formation: lighting and geometry. As such, we propose a novel
evaluation measure, parametric norm-balls, by directly perturbing physical
parameters that underly image formation. One enabling contribution we present
is a physically-based differentiable renderer that allows us to propagate pixel
gradients to the parametric space of lighting and geometry. Our approach
enables physically-based adversarial attacks, and our differentiable renderer
leverages models from the interactive rendering literature to balance the
performance and accuracy trade-offs necessary for a memory-efficient and
scalable adversarial data augmentation workflow.
|
Can Network Analysis Techniques help to Predict Design Dependencies? An
Initial Study
|
The degree of dependencies among the modules of a software system is a key
attribute to characterize its design structure and its ability to evolve over
time. Several design problems are often correlated with undesired dependencies
among modules. Being able to anticipate those problems is important for
developers, so they can plan early for maintenance and refactoring efforts.
However, existing tools are limited to detecting undesired dependencies once
they appeared in the system. In this work, we investigate whether module
dependencies can be predicted (before they actually appear). Since the module
structure can be regarded as a network, i.e, a dependency graph, we leverage on
network features to analyze the dynamics of such a structure. In particular, we
apply link prediction techniques for this task. We conducted an evaluation on
two Java projects across several versions, using link prediction and machine
learning techniques, and assessed their performance for identifying new
dependencies from a project version to the next one. The results, although
preliminary, show that the link prediction approach is feasible for package
dependencies. Also, this work opens opportunities for further development of
software-specific strategies for dependency prediction.
|
Highly Accelerated Multishot EPI through Synergistic Machine Learning
and Joint Reconstruction
|
Purpose: To introduce a combined machine learning (ML) and physics-based
image reconstruction framework that enables navigator-free, highly accelerated
multishot echo planar imaging (msEPI), and demonstrate its application in
high-resolution structural and diffusion imaging.
Methods: Singleshot EPI is an efficient encoding technique, but does not lend
itself well to high-resolution imaging due to severe distortion artifacts and
blurring. While msEPI can mitigate these artifacts, high-quality msEPI has been
elusive because of phase mismatch arising from shot-to-shot variations which
preclude the combination of the multiple-shot data into a single image. We
employ deep learning to obtain an interim image with minimal artifacts, which
permits estimation of image phase variations due to shot-to-shot changes. These
variations are then included in a Joint Virtual Coil Sensitivity Encoding
(JVC-SENSE) reconstruction to utilize data from all shots and improve upon the
ML solution.
Results: Our combined ML + physics approach enabled Rinplane x MultiBand (MB)
= 8x2-fold acceleration using 2 EPI-shots for multi-echo imaging, so that
whole-brain T2 and T2* parameter maps could be derived from an 8.3 sec
acquisition at 1x1x3mm3 resolution. This has also allowed high-resolution
diffusion imaging with high geometric fidelity using 5-shots at Rinplane x MB =
9x2-fold acceleration. To make these possible, we extended the state-of-the-art
MUSSELS reconstruction technique to Simultaneous MultiSlice (SMS) encoding and
used it as an input to our ML network.
Conclusion: Combination of ML and JVC-SENSE enabled navigator-free msEPI at
higher accelerations than previously possible while using fewer shots, with
reduced vulnerability to poor generalizability and poor acceptance of
end-to-end ML approaches.
|
Backprop Evolution
|
The back-propagation algorithm is the cornerstone of deep learning. Despite
its importance, few variations of the algorithm have been attempted. This work
presents an approach to discover new variations of the back-propagation
equation. We use a domain specific lan- guage to describe update equations as a
list of primitive functions. An evolution-based method is used to discover new
propagation rules that maximize the generalization per- formance after a few
epochs of training. We find several update equations that can train faster with
short training times than standard back-propagation, and perform similar as
standard back-propagation at convergence.
|
On the Effect of Task-to-Worker Assignment in Distributed Computing
Systems with Stragglers
|
We study the expected completion time of some recently proposed algorithms
for distributed computing which redundantly assign computing tasks to multiple
machines in order to tolerate a certain number of machine failures. We
analytically show that not only the amount of redundancy but also the
task-to-machine assignments affect the latency in a distributed system. We
study systems with a fixed number of computing tasks that are split in possibly
overlapping batches, and independent exponentially distributed machine service
times. We show that, for such systems, the uniform replication of non-
overlapping (disjoint) batches of computing tasks achieves the minimum expected
computing time.
|
Random directions stochastic approximation with deterministic
perturbations
|
We introduce deterministic perturbation schemes for the recently proposed
random directions stochastic approximation (RDSA) [17], and propose new
first-order and second-order algorithms. In the latter case, these are the
first second-order algorithms to incorporate deterministic perturbations. We
show that the gradient and/or Hessian estimates in the resulting algorithms
with deterministic perturbations are asymptotically unbiased, so that the
algorithms are provably convergent. Furthermore, we derive convergence rates to
establish the superiority of the first-order and second-order algorithms, for
the special case of a convex and quadratic optimization problem, respectively.
Numerical experiments are used to validate the theoretical results.
|
Nonparametric Gaussian Mixture Models for the Multi-Armed Contextual
Bandit
|
We here adopt Bayesian nonparametric mixture models to extend multi-armed
bandits in general, and Thompson sampling in particular, to scenarios where
there is reward model uncertainty. In the stochastic multi-armed bandit, where
an agent must learn a policy that maximizes long term payoff, the reward for
the selected action is generated from an unknown distribution. Thompson
sampling is a generative and interpretable multi-armed bandit algorithm that
has been shown both to perform well in practice, and to enjoy optimality
properties for certain reward functions. Nevertheless, Thompson sampling
requires knowledge of the true reward model, for calculation of expected
rewards and sampling from its parameter posterior. In this work, we extend
Thompson sampling to complex scenarios where there is model uncertainty, by
adopting a very flexible set of reward distributions: Bayesian nonparametric
Gaussian mixture models. The generative process of Bayesian nonparametric
mixtures naturally aligns with the Bayesian modeling of multi-armed bandits:
the nonparametric model autonomously determines its complexity as new rewards
are observed for the played arms. By characterizing each arm's reward
distribution with independent nonparametric mixture models, the proposed method
sequentially learns the model that best approximates the true underlying reward
distribution, achieving successful performance in complex -- not in the
exponential family -- bandits. Our contribution is valuable for practical
scenarios, as it avoids stringent case-by-case model specifications and
hyperparameter tuning, yet attains reduced regret in diverse bandit settings.
|
(Sequential) Importance Sampling Bandits
|
This work extends existing multi-armed bandit (MAB) algorithms beyond their
original settings by leveraging advances in sequential Monte Carlo (SMC)
methods from the approximate inference community. We leverage Monte Carlo
estimation and, in particular, the flexibility of (sequential) importance
sampling to allow for accurate estimation of the statistics of interest within
the MAB problem. The MAB is a sequential allocation task where the goal is to
learn a policy that maximizes long term payoff, where only the reward of the
executed action is observed; i.e., sequential optimal decisions are made, while
simultaneously learning how the world operates. In the stochastic setting, the
reward for each action is generated from an unknown distribution. To decide the
next optimal action to take, one must compute sufficient statistics of this
unknown reward distribution, e.g., upper-confidence bounds (UCB), or
expectations in Thompson sampling. Closed-form expressions for these statistics
of interest are analytically intractable except for simple cases. By combining
SMC methods --- which estimate posterior densities and expectations in
probabilistic models that are analytically intractable --- with Bayesian
state-of-the-art MAB algorithms, we extend their applicability to complex
models: those for which sampling may be performed even if analytic computation
of summary statistics is infeasible --- nonlinear reward functions and dynamic
bandits. We combine SMC both for Thompson sampling and upper confident
bound-based (Bayes-UCB) policies, and study different bandit models: classic
Bernoulli and Gaussian distributed cases, as well as dynamic and context
dependent linear-Gaussian, logistic and categorical-softmax rewards.
|
Towards Learning Fine-Grained Disentangled Representations from Speech
|
Learning disentangled representations of high-dimensional data is currently
an active research area. However, compared to the field of computer vision,
less work has been done for speech processing. In this paper, we provide a
review of two representative efforts on this topic and propose the novel
concept of fine-grained disentangled speech representation learning.
|
On the Convergence of A Class of Adam-Type Algorithms for Non-Convex
Optimization
|
This paper studies a class of adaptive gradient based momentum algorithms
that update the search directions and learning rates simultaneously using past
gradients. This class, which we refer to as the "Adam-type", includes the
popular algorithms such as the Adam, AMSGrad and AdaGrad. Despite their
popularity in training deep neural networks, the convergence of these
algorithms for solving nonconvex problems remains an open question. This paper
provides a set of mild sufficient conditions that guarantee the convergence for
the Adam-type methods. We prove that under our derived conditions, these
methods can achieve the convergence rate of order $O(\log{T}/\sqrt{T})$ for
nonconvex stochastic optimization. We show the conditions are essential in the
sense that violating them may make the algorithm diverge. Moreover, we propose
and analyze a class of (deterministic) incremental adaptive gradient
algorithms, which has the same $O(\log{T}/\sqrt{T})$ convergence rate. Our
study could also be extended to a broader class of adaptive gradient methods in
machine learning and optimization.
|
Feature Dimensionality Reduction for Video Affect Classification: A
Comparative Study
|
Affective computing has become a very important research area in
human-machine interaction. However, affects are subjective, subtle, and
uncertain. So, it is very difficult to obtain a large number of labeled
training samples, compared with the number of possible features we could
extract. Thus, dimensionality reduction is critical in affective computing.
This paper presents our preliminary study on dimensionality reduction for
affect classification. Five popular dimensionality reduction approaches are
introduced and compared. Experiments on the DEAP dataset showed that no
approach can universally outperform others, and performing classification using
the raw features directly may not always be a bad choice.
|
Compressed Sensing Using Binary Matrices of Nearly Optimal Dimensions
|
In this paper, we study the problem of compressed sensing using binary
measurement matrices and $\ell_1$-norm minimization (basis pursuit) as the
recovery algorithm. We derive new upper and lower bounds on the number of
measurements to achieve robust sparse recovery with binary matrices. We
establish sufficient conditions for a column-regular binary matrix to satisfy
the robust null space property (RNSP) and show that the associated sufficient
conditions % sparsity bounds for robust sparse recovery obtained using the RNSP
are better by a factor of $(3 \sqrt{3})/2 \approx 2.6$ compared to the
sufficient conditions obtained using the restricted isometry property (RIP).
Next we derive universal \textit{lower} bounds on the number of measurements
that any binary matrix needs to have in order to satisfy the weaker sufficient
condition based on the RNSP and show that bipartite graphs of girth six are
optimal. Then we display two classes of binary matrices, namely parity check
matrices of array codes and Euler squares, which have girth six and are nearly
optimal in the sense of almost satisfying the lower bound. In principle,
randomly generated Gaussian measurement matrices are "order-optimal". So we
compare the phase transition behavior of the basis pursuit formulation using
binary array codes and Gaussian matrices and show that (i) there is essentially
no difference between the phase transition boundaries in the two cases and (ii)
the CPU time of basis pursuit with binary matrices is hundreds of times faster
than with Gaussian matrices and the storage requirements are less. Therefore it
is suggested that binary matrices are a viable alternative to Gaussian matrices
for compressed sensing using basis pursuit. \end{abstract}
|
Policy Optimization as Wasserstein Gradient Flows
|
Policy optimization is a core component of reinforcement learning (RL), and
most existing RL methods directly optimize parameters of a policy based on
maximizing the expected total reward, or its surrogate. Though often achieving
encouraging empirical success, its underlying mathematical principle on {\em
policy-distribution} optimization is unclear. We place policy optimization into
the space of probability measures, and interpret it as Wasserstein gradient
flows. On the probability-measure space, under specified circumstances, policy
optimization becomes a convex problem in terms of distribution optimization. To
make optimization feasible, we develop efficient algorithms by numerically
solving the corresponding discrete gradient flows. Our technique is applicable
to several RL settings, and is related to many state-of-the-art
policy-optimization algorithms. Empirical results verify the effectiveness of
our framework, often obtaining better performance compared to related
algorithms.
|
Gradient and Newton Boosting for Classification and Regression
|
Boosting algorithms are frequently used in applied data science and in
research. To date, the distinction between boosting with either gradient
descent or second-order Newton updates is often not made in both applied and
methodological research, and it is thus implicitly assumed that the difference
is irrelevant. The goal of this article is to clarify this situation. In
particular, we present gradient and Newton boosting, as well as a hybrid
variant of the two, in a unified framework. We compare these boosting
algorithms with trees as base learners using various datasets and loss
functions. Our experiments show that Newton boosting outperforms gradient and
hybrid gradient-Newton boosting in terms of predictive accuracy on the majority
of datasets. We also present evidence that the reason for this is not faster
convergence of Newton boosting. In addition, we introduce a novel tuning
parameter for tree-based Newton boosting which is interpretable and important
for predictive accuracy.
|
On feature selection and evaluation of transportation mode prediction
strategies
|
Transportation modes prediction is a fundamental task for decision making in
smart cities and traffic management systems. Traffic policies designed based on
trajectory mining can save money and time for authorities and the public. It
may reduce the fuel consumption and commute time and moreover, may provide more
pleasant moments for residents and tourists. Since the number of features that
may be used to predict a user transportation mode can be substantial, finding a
subset of features that maximizes a performance measure is worth investigating.
In this work, we explore wrapper and information retrieval methods to find the
best subset of trajectory features. After finding the best classifier and the
best feature subset, our results were compared with two related papers that
applied deep learning methods and the results showed that our framework
achieved better performance. Furthermore, two types of cross-validation
approaches were investigated, and the performance results show that the random
cross-validation method provides optimistic results.
|
Classifier-Guided Visual Correction of Noisy Labels for Image
Classification Tasks
|
Training data plays an essential role in modern applications of machine
learning. However, gathering labeled training data is time-consuming.
Therefore, labeling is often outsourced to less experienced users, or
completely automated. This can introduce errors, which compromise valuable
training data, and lead to suboptimal training results. We thus propose a novel
approach that uses the power of pretrained classifiers to visually guide users
to noisy labels, and let them interactively check error candidates, to
iteratively improve the training data set. To systematically investigate
training data, we propose a categorization of labeling errors into three
different types, based on an analysis of potential pitfalls in label
acquisition processes. For each of these types, we present approaches to
detect, reason about, and resolve error candidates, as we propose measures and
visual guidance techniques to support machine learning users. Our approach has
been used to spot errors in well-known machine learning benchmark data sets,
and we tested its usability during a user evaluation. While initially developed
for images, the techniques presented in this paper are independent of the
classification algorithm, and can also be extended to many other types of
training data.
|
A New Optimization Layer for Real-Time Bidding Advertising Campaigns
|
While it is relatively easy to start an online advertising campaign,
obtaining a high Key Performance Indicator (KPI) can be challenging. A large
body of work on this subject has already been performed and platforms known as
DSPs are available on the market that deal with such an optimization. From the
advertiser's point of view, each DSP is a different black box, with its pros
and cons, that needs to be configured. In order to take advantage of the pros
of every DSP, advertisers are well-advised to use a combination of them when
setting up their campaigns. In this paper, we propose an algorithm for
advertisers to add an optimization layer on top of DSPs. The algorithm we
introduce, called SKOTT, maximizes the chosen KPI by optimally configuring the
DSPs and putting them in competition with each other. SKOTT is a highly
specialized iterative algorithm loosely based on gradient descent that is made
up of three independent sub-routines, each dealing with a different problem:
partitioning the budget, setting the desired average bid, and preventing
under-delivery. In particular, one of the novelties of our approach lies in our
taking the perspective of the advertisers rather than the DSPs. Synthetic
market data is used to evaluate the efficiency of SKOTT against other
state-of-the-art approaches adapted from similar problems. The results
illustrate the benefits of our proposals, which greatly outperforms the other
methods.
|
Data-driven polynomial chaos expansion for machine learning regression
|
We present a regression technique for data-driven problems based on
polynomial chaos expansion (PCE). PCE is a popular technique in the field of
uncertainty quantification (UQ), where it is typically used to replace a
runnable but expensive computational model subject to random inputs with an
inexpensive-to-evaluate polynomial function. The metamodel obtained enables a
reliable estimation of the statistics of the output, provided that a suitable
probabilistic model of the input is available. Machine learning (ML) regression
is a research field that focuses on providing purely data-driven input-output
maps, with the focus on pointwise prediction accuracy. We show that a PCE
metamodel purely trained on data can yield pointwise predictions whose accuracy
is comparable to that of other ML regression models, such as neural networks
and support vector machines. The comparisons are performed on benchmark
datasets available from the literature. The methodology also enables the
quantification of the output uncertainties, and is robust to noise.
Furthermore, it enjoys additional desirable properties, such as good
performance for small training sets and simplicity of construction, with only
little parameter tuning required.
|
Identifying Protein-Protein Interaction using Tree LSTM and Structured
Attention
|
Identifying interactions between proteins is important to understand
underlying biological processes. Extracting a protein-protein interaction (PPI)
from the raw text is often very difficult. Previous supervised learning methods
have used handcrafted features on human-annotated data sets. In this paper, we
propose a novel tree recurrent neural network with structured attention
architecture for doing PPI. Our architecture achieves state of the art results
(precision, recall, and F1-score) on the AIMed and BioInfer benchmark data
sets. Moreover, our models achieve a significant improvement over previous best
models without any explicit feature extraction. Our experimental results show
that traditional recurrent networks have inferior performance compared to tree
recurrent networks for the supervised PPI problem.
|
Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal
densities?
|
Hamiltonian Monte Carlo (HMC) is a very popular and generic collection of
Markov chain Monte Carlo (MCMC) algorithms. One explanation for the popularity
of HMC algorithms is their excellent performance as the dimension $d$ of the
target becomes large: under conditions that are satisfied for many common
statistical models, optimally-tuned HMC algorithms have a running time that
scales like $d^{0.25}$. In stark contrast, the running time of the usual
Random-Walk Metropolis (RWM) algorithm, optimally tuned, scales like $d$. This
superior scaling of the HMC algorithm with dimension is attributed to the fact
that it, unlike RWM, incorporates the gradient information in the proposal
distribution. In this paper, we investigate a different scaling question: does
HMC beat RWM for highly $\textit{multimodal}$ targets? We find that the answer
is often $\textit{no}$. We compute the spectral gaps for both the algorithms
for a specific class of multimodal target densities, and show that they are
identical. The key reason is that, within one mode, the gradient is effectively
ignorant about other modes, thus negating the advantage the HMC algorithm
enjoys in unimodal targets. We also give heuristic arguments suggesting that
the above observation may hold quite generally. Our main tool for answering
this question is a novel simple formula for the conductance of HMC using
Liouville's theorem. This result allows us to compute the spectral gap of HMC
algorithms, for both the classical HMC with isotropic momentum and the recent
Riemannian HMC, for multimodal targets.
|
OBOE: Collaborative Filtering for AutoML Model Selection
|
Algorithm selection and hyperparameter tuning remain two of the most
challenging tasks in machine learning. Automated machine learning (AutoML)
seeks to automate these tasks to enable widespread use of machine learning by
non-experts. This paper introduces OBOE, a collaborative filtering method for
time-constrained model selection and hyperparameter tuning. OBOE forms a matrix
of the cross-validated errors of a large number of supervised learning models
(algorithms together with hyperparameters) on a large number of datasets, and
fits a low rank model to learn the low-dimensional feature vectors for the
models and datasets that best predict the cross-validated errors. To find
promising models for a new dataset, OBOE runs a set of fast but informative
algorithms on the new dataset and uses their cross-validated errors to infer
the feature vector for the new dataset. OBOE can find good models under
constraints on the number of models fit or the total time budget. To this end,
this paper develops a new heuristic for active learning in time-constrained
matrix completion based on optimal experiment design. Our experiments
demonstrate that OBOE delivers state-of-the-art performance faster than
competing approaches on a test bed of supervised learning problems. Moreover,
the success of the bilinear model used by OBOE suggests that AutoML may be
simpler than was previously understood.
|
Augmenting Physical Simulators with Stochastic Neural Networks: Case
Study of Planar Pushing and Bouncing
|
An efficient, generalizable physical simulator with universal uncertainty
estimates has wide applications in robot state estimation, planning, and
control. In this paper, we build such a simulator for two scenarios, planar
pushing and ball bouncing, by augmenting an analytical rigid-body simulator
with a neural network that learns to model uncertainty as residuals. Combining
symbolic, deterministic simulators with learnable, stochastic neural nets
provides us with expressiveness, efficiency, and generalizability
simultaneously. Our model outperforms both purely analytical and purely learned
simulators consistently on real, standard benchmarks. Compared with methods
that model uncertainty using Gaussian processes, our model runs much faster,
generalizes better to new object shapes, and is able to characterize the
complex distribution of object trajectories.
|
Counterfactual Normalization: Proactively Addressing Dataset Shift and
Improving Reliability Using Causal Mechanisms
|
Predictive models can fail to generalize from training to deployment
environments because of dataset shift, posing a threat to model reliability and
the safety of downstream decisions made in practice. Instead of using samples
from the target distribution to reactively correct dataset shift, we use
graphical knowledge of the causal mechanisms relating variables in a prediction
problem to proactively remove relationships that do not generalize across
environments, even when these relationships may depend on unobserved variables
(violations of the "no unobserved confounders" assumption). To accomplish this,
we identify variables with unstable paths of statistical influence and remove
them from the model. We also augment the causal graph with latent
counterfactual variables that isolate unstable paths of statistical influence,
allowing us to retain stable paths that would otherwise be removed. Our
experiments demonstrate that models that remove vulnerable variables and use
estimates of the latent variables transfer better, often outperforming in the
target domain despite some accuracy loss in the training domain.
|
Application of Bounded Total Variation Denoising in Urban Traffic
Analysis
|
While it is believed that denoising is not always necessary in many big data
applications, we show in this paper that denoising is helpful in urban traffic
analysis by applying the method of bounded total variation denoising to the
urban road traffic prediction and clustering problem. We propose two
easy-to-implement methods to estimate the noise strength parameter in the
denoising algorithm, and apply the denoising algorithm to GPS-based traffic
data from Beijing taxi system. For the traffic prediction problem, we combine
neural network and history matching method for roads randomly chosen from an
urban area of Beijing. Numerical experiments show that the predicting accuracy
is improved significantly by applying the proposed bounded total variation
denoising algorithm. We also test the algorithm on clustering problem, where a
recently developed clustering analysis method is applied to more than one
hundred urban road segments in Beijing based on their velocity profiles. Better
clustering result is obtained after denoising.
|
A Hybrid Recommender System for Patient-Doctor Matchmaking in Primary
Care
|
We partner with a leading European healthcare provider and design a mechanism
to match patients with family doctors in primary care. We define the
matchmaking process for several distinct use cases given different levels of
available information about patients. Then, we adopt a hybrid recommender
system to present each patient a list of family doctor recommendations. In
particular, we model patient trust of family doctors using a large-scale
dataset of consultation histories, while accounting for the temporal dynamics
of their relationships. Our proposed approach shows higher predictive accuracy
than both a heuristic baseline and a collaborative filtering approach, and the
proposed trust measure further improves model performance.
|
VerIDeep: Verifying Integrity of Deep Neural Networks through
Sensitive-Sample Fingerprinting
|
Deep learning has become popular, and numerous cloud-based services are
provided to help customers develop and deploy deep learning applications.
Meanwhile, various attack techniques have also been discovered to stealthily
compromise the model's integrity. When a cloud customer deploys a deep learning
model in the cloud and serves it to end-users, it is important for him to be
able to verify that the deployed model has not been tampered with, and the
model's integrity is protected.
We propose a new low-cost and self-served methodology for customers to verify
that the model deployed in the cloud is intact, while having only black-box
access (e.g., via APIs) to the deployed model. Customers can detect arbitrary
changes to their deep learning models. Specifically, we define
\texttt{Sensitive-Sample} fingerprints, which are a small set of transformed
inputs that make the model outputs sensitive to the model's parameters. Even
small weight changes can be clearly reflected in the model outputs, and
observed by the customer. Our experiments on different types of model integrity
attacks show that we can detect model integrity breaches with high accuracy
($>$99\%) and low overhead ($<$10 black-box model accesses).
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.