title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Uncertainty in the Variational Information Bottleneck
|
We present a simple case study, demonstrating that Variational Information
Bottleneck (VIB) can improve a network's classification calibration as well as
its ability to detect out-of-distribution data. Without explicitly being
designed to do so, VIB gives two natural metrics for handling and quantifying
uncertainty.
|
Semantic Segmentation with Scarce Data
|
Semantic segmentation is a challenging vision problem that usually
necessitates the collection of large amounts of finely annotated data, which is
often quite expensive to obtain. Coarsely annotated data provides an
interesting alternative as it is usually substantially more cheap. In this
work, we present a method to leverage coarsely annotated data along with fine
supervision to produce better segmentation results than would be obtained when
training using only the fine data. We validate our approach by simulating a
scarce data setting with less than 200 low resolution images from the
Cityscapes dataset and show that our method substantially outperforms solely
training on the fine annotation data by an average of 15.52% mIoU and
outperforms the coarse mask by an average of 5.28% mIoU.
|
Mining Illegal Insider Trading of Stocks: A Proactive Approach
|
Illegal insider trading of stocks is based on releasing non-public
information (e.g., new product launch, quarterly financial report, acquisition
or merger plan) before the information is made public. Detecting illegal
insider trading is difficult due to the complex, nonlinear, and non-stationary
nature of the stock market. In this work, we present an approach that detects
and predicts illegal insider trading proactively from large heterogeneous
sources of structured and unstructured data using a deep-learning based
approach combined with discrete signal processing on the time series data. In
addition, we use a tree-based approach that visualizes events and actions to
aid analysts in their understanding of large amounts of unstructured data.
Using existing data, we have discovered that our approach has a good success
rate in detecting illegal insider trading patterns.
|
Stochastic Layer-Wise Precision in Deep Neural Networks
|
Low precision weights, activations, and gradients have been proposed as a way
to improve the computational efficiency and memory footprint of deep neural
networks. Recently, low precision networks have even shown to be more robust to
adversarial attacks. However, typical implementations of low precision DNNs use
uniform precision across all layers of the network. In this work, we explore
whether a heterogeneous allocation of precision across a network leads to
improved performance, and introduce a learning scheme where a DNN
stochastically explores multiple precision configurations through learning.
This permits a network to learn an optimal precision configuration. We show on
convolutional neural networks trained on MNIST and ILSVRC12 that even though
these nets learn a uniform or near-uniform allocation strategy respectively,
stochastic precision leads to a favourable regularization effect improving
generalization.
|
Structure Learning of Markov Random Fields through Grow-Shrink Maximum
Pseudolikelihood Estimation
|
Learning the structure of Markov random fields (MRFs) plays an important role
in multivariate analysis. The importance has been increasing with the recent
rise of statistical relational models since the MRF serves as a building block
of these models such as Markov logic networks. There are two fundamental ways
to learn structures of MRFs: methods based on parameter learning and those
based on independence test. The former methods more or less assume certain
forms of distribution, so they potentially perform poorly when the assumption
is not satisfied. The latter can learn an MRF structure without a strong
distributional assumption, but sometimes it is unclear what objective function
is maximized/minimized in these methods. In this paper, we follow the latter,
but we explicitly define the optimization problem of MRF structure learning as
maximum pseudolikelihood estimation (MPLE) with respect to the edge set. As a
result, the proposed solution successfully deals with the {\em symmetricity} in
MRFs, whereas such symmetricity is not taken into account in most existing
independence test techniques. The proposed method achieved higher accuracy than
previous methods when there were asymmetric dependencies in our experiments.
|
Scalable Structure Learning for Probabilistic Soft Logic
|
Statistical relational frameworks such as Markov logic networks and
probabilistic soft logic (PSL) encode model structure with weighted first-order
logical clauses. Learning these clauses from data is referred to as structure
learning. Structure learning alleviates the manual cost of specifying models.
However, this benefit comes with high computational costs; structure learning
typically requires an expensive search over the space of clauses which involves
repeated optimization of clause weights. In this paper, we propose the first
two approaches to structure learning for PSL. We introduce a greedy
search-based algorithm and a novel optimization method that trade-off
scalability and approximations to the structure learning problem in varying
ways. The highly scalable optimization method combines data-driven generation
of clauses with a piecewise pseudolikelihood (PPLL) objective that learns model
structure by optimizing clause weights only once. We compare both methods
across five real-world tasks, showing that PPLL achieves an order of magnitude
runtime speedup and AUC gains up to 15% over greedy search.
|
Modular Vehicle Control for Transferring Semantic Information Between
Weather Conditions Using GANs
|
Even though end-to-end supervised learning has shown promising results for
sensorimotor control of self-driving cars, its performance is greatly affected
by the weather conditions under which it was trained, showing poor
generalization to unseen conditions. In this paper, we show how knowledge can
be transferred using semantic maps to new weather conditions without the need
to obtain new ground truth data. To this end, we propose to divide the task of
vehicle control into two independent modules: a control module which is only
trained on one weather condition for which labeled steering data is available,
and a perception module which is used as an interface between new weather
conditions and the fixed control module. To generate the semantic data needed
to train the perception module, we propose to use a generative adversarial
network (GAN)-based model to retrieve the semantic information for the new
conditions in an unsupervised manner. We introduce a master-servant
architecture, where the master model (semantic labels available) trains the
servant model (semantic labels not available). We show that our proposed method
trained with ground truth data for a single weather condition is capable of
achieving similar results on the task of steering angle prediction as an
end-to-end model trained with ground truth data of 15 different weather
conditions.
|
Coopetitive Soft Gating Ensemble
|
In this article, we propose the Coopetititve Soft Gating Ensemble or CSGE for
general machine learning tasks and interwoven systems. The goal of machine
learning is to create models that generalize well for unknown datasets. Often,
however, the problems are too complex to be solved with a single model, so
several models are combined. Similar, Autonomic Computing requires the
integration of different systems. Here, especially, the local, temporal online
evaluation and the resulting (re-)weighting scheme of the CSGE makes the
approach highly applicable for self-improving system integrations. To achieve
the best potential performance the CSGE can be optimized according to arbitrary
loss functions making it accessible for a broader range of problems. We
introduce a novel training procedure including a hyper-parameter initialisation
at its heart. We show that the CSGE approach reaches state-of-the-art
performance for both classification and regression tasks. Further on, the CSGE
provides a human-readable quantification on the influence of all base
estimators employing the three weighting aspects. Moreover, we provide a
scikit-learn compatible implementation.
|
When Gaussian Process Meets Big Data: A Review of Scalable GPs
|
The vast quantity of information brought by big data as well as the evolving
computer hardware encourages success stories in the machine learning community.
In the meanwhile, it poses challenges for the Gaussian process (GP) regression,
a well-known non-parametric and interpretable Bayesian model, which suffers
from cubic complexity to data size. To improve the scalability while retaining
desirable prediction quality, a variety of scalable GPs have been presented.
But they have not yet been comprehensively reviewed and analyzed in order to be
well understood by both academia and industry. The review of scalable GPs in
the GP community is timely and important due to the explosion of data size. To
this end, this paper is devoted to the review on state-of-the-art scalable GPs
involving two main categories: global approximations which distillate the
entire data and local approximations which divide the data for subspace
learning. Particularly, for global approximations, we mainly focus on sparse
approximations comprising prior approximations which modify the prior but
perform exact inference, posterior approximations which retain exact prior but
perform approximate inference, and structured sparse approximations which
exploit specific structures in kernel matrix; for local approximations, we
highlight the mixture/product of experts that conducts model averaging from
multiple local experts to boost predictions. To present a complete review,
recent advances for improving the scalability and capability of scalable GPs
are reviewed. Finally, the extensions and open issues regarding the
implementation of scalable GPs in various scenarios are reviewed and discussed
to inspire novel ideas for future research avenues.
|
Behaviour Policy Estimation in Off-Policy Policy Evaluation: Calibration
Matters
|
In this work, we consider the problem of estimating a behaviour policy for
use in Off-Policy Policy Evaluation (OPE) when the true behaviour policy is
unknown. Via a series of empirical studies, we demonstrate how accurate OPE is
strongly dependent on the calibration of estimated behaviour policy models: how
precisely the behaviour policy is estimated from data. We show how powerful
parametric models such as neural networks can result in highly uncalibrated
behaviour policy models on a real-world medical dataset, and illustrate how a
simple, non-parametric, k-nearest neighbours model produces better calibrated
behaviour policy estimates and can be used to obtain superior importance
sampling-based OPE estimates.
|
Adversarial Robustness Toolbox v1.0.0
|
Adversarial Robustness Toolbox (ART) is a Python library supporting
developers and researchers in defending Machine Learning models (Deep Neural
Networks, Gradient Boosted Decision Trees, Support Vector Machines, Random
Forests, Logistic Regression, Gaussian Processes, Decision Trees, Scikit-learn
Pipelines, etc.) against adversarial threats and helps making AI systems more
secure and trustworthy. Machine Learning models are vulnerable to adversarial
examples, which are inputs (images, texts, tabular data, etc.) deliberately
modified to produce a desired response by the Machine Learning model. ART
provides the tools to build and deploy defences and test them with adversarial
attacks. Defending Machine Learning models involves certifying and verifying
model robustness and model hardening with approaches such as pre-processing
inputs, augmenting training data with adversarial samples, and leveraging
runtime detection methods to flag any inputs that might have been modified by
an adversary. The attacks implemented in ART allow creating adversarial attacks
against Machine Learning models which is required to test defenses with
state-of-the-art threat models. Supported Machine Learning Libraries include
TensorFlow (v1 and v2), Keras, PyTorch, MXNet, Scikit-learn, XGBoost, LightGBM,
CatBoost, and GPy. The source code of ART is released with MIT license at
https://github.com/IBM/adversarial-robustness-toolbox. The release includes
code examples, notebooks with tutorials and documentation
(http://adversarial-robustness-toolbox.readthedocs.io).
|
Domain Aware Markov Logic Networks
|
Combining logic and probability has been a long stand- ing goal of AI
research. Markov Logic Networks (MLNs) achieve this by attaching weights to
formulas in first-order logic, and can be seen as templates for constructing
features for ground Markov networks. Most techniques for learning weights of
MLNs are domain-size agnostic, i.e., the size of the domain is not explicitly
taken into account while learn- ing the parameters of the model. This often
results in ex- treme probabilities when testing on domain sizes different from
those seen during training. In this paper, we propose Domain Aware Markov logic
Networks (DA-MLNs) which present a principled solution to this problem. While
defin- ing the ground network distribution, DA-MLNs divide the ground feature
weight by a scaling factor which is a function of the number of connections the
ground atoms appearing in the feature are involved in. We show that standard
MLNs fall out as a special case of our formalism when this func- tion evaluates
to a constant equal to 1. Experiments on the benchmark Friends & Smokers domain
show that our ap- proach results in significantly higher accuracies compared to
existing methods when testing on domains whose sizes different from those seen
during training.
|
A Mean-Field Optimal Control Formulation of Deep Learning
|
Recent work linking deep neural networks and dynamical systems opened up new
avenues to analyze deep learning. In particular, it is observed that new
insights can be obtained by recasting deep learning as an optimal control
problem on difference or differential equations. However, the mathematical
aspects of such a formulation have not been systematically explored. This paper
introduces the mathematical formulation of the population risk minimization
problem in deep learning as a mean-field optimal control problem. Mirroring the
development of classical optimal control, we state and prove optimality
conditions of both the Hamilton-Jacobi-Bellman type and the Pontryagin type.
These mean-field results reflect the probabilistic nature of the learning
problem. In addition, by appealing to the mean-field Pontryagin's maximum
principle, we establish some quantitative relationships between population and
empirical learning problems. This serves to establish a mathematical foundation
for investigating the algorithmic and theoretical connections between optimal
control and deep learning.
|
One-Class Kernel Spectral Regression
|
The paper introduces a new efficient nonlinear one-class classifier
formulated as the Rayleigh quotient criterion optimisation. The method,
operating in a reproducing kernel Hilbert space, minimises the scatter of
target distribution along an optimal projection direction while at the same
time keeping projections of positive observations distant from the mean of the
negative class. We provide a graph embedding view of the problem which can then
be solved efficiently using the spectral regression approach. In this sense,
unlike previous similar methods which often require costly eigen-computations
of dense matrices, the proposed approach casts the problem under consideration
into a regression framework which is computationally more efficient. In
particular, it is shown that the dominant complexity of the proposed method is
the complexity of computing the kernel matrix. Additional appealing
characteristics of the proposed one-class classifier are: 1-the ability to be
trained in an incremental fashion (allowing for application in streaming data
scenarios while also reducing the computational complexity in a non-streaming
operation mode); 2-being unsupervised, but providing the option for refining
the solution using negative training examples, when available; And last but not
the least, 3-the use of the kernel trick which facilitates a nonlinear mapping
of the data into a high-dimensional feature space to seek better solutions.
|
Weakly Supervised Deep Recurrent Neural Networks for Basic Dance Step
Generation
|
Synthesizing human's movements such as dancing is a flourishing research
field which has several applications in computer graphics. Recent studies have
demonstrated the advantages of deep neural networks (DNNs) for achieving
remarkable performance in motion and music tasks with little effort for feature
pre-processing. However, applying DNNs for generating dance to a piece of music
is nevertheless challenging, because of 1) DNNs need to generate large
sequences while mapping the music input, 2) the DNN needs to constraint the
motion beat to the music, and 3) DNNs require a considerable amount of
hand-crafted data. In this study, we propose a weakly supervised deep recurrent
method for real-time basic dance generation with audio power spectrum as input.
The proposed model employs convolutional layers and a multilayered Long
Short-Term memory (LSTM) to process the audio input. Then, another deep LSTM
layer decodes the target dance sequence. Notably, this end-to-end approach has
1) an auto-conditioned decode configuration that reduces accumulation of
feedback error of large dance sequence, 2) uses a contrastive cost function to
regulate the mapping between the music and motion beat, and 3) trains with weak
labels generated from the motion beat, reducing the amount of hand-crafted
data. We evaluate the proposed network based on i) the similarities between
generated and the baseline dancer motion with a cross entropy measure for large
dance sequences, and ii) accurate timing between the music and motion beat with
an F-measure. Experimental results revealed that, after training using a small
dataset, the model generates basic dance steps with low cross entropy and
maintains an F-measure score similar to that of a baseline dancer.
|
Welfare and Distributional Impacts of Fair Classification
|
Current methodologies in machine learning analyze the effects of various
statistical parity notions of fairness primarily in light of their impacts on
predictive accuracy and vendor utility loss. In this paper, we propose a new
framework for interpreting the effects of fairness criteria by converting the
constrained loss minimization problem into a social welfare maximization
problem. This translation moves a classifier and its output into utility space
where individuals, groups, and society at-large experience different welfare
changes due to classification assignments. Under this characterization,
predictions and fairness constraints are seen as shaping societal welfare and
distribution and revealing individuals' implied welfare weights in
society--weights that may then be interpreted through a fairness lens. The
social welfare formulation of the fairness problem brings to the fore concerns
of distributive justice that have always had a central albeit more implicit
role in standard algorithmic fairness approaches.
|
Credit Default Mining Using Combined Machine Learning and Heuristic
Approach
|
Predicting potential credit default accounts in advance is challenging.
Traditional statistical techniques typically cannot handle large amounts of
data and the dynamic nature of fraud and humans. To tackle this problem, recent
research has focused on artificial and computational intelligence based
approaches. In this work, we present and validate a heuristic approach to mine
potential default accounts in advance where a risk probability is precomputed
from all previous data and the risk probability for recent transactions are
computed as soon they happen. Beside our heuristic approach, we also apply a
recently proposed machine learning approach that has not been applied
previously on our targeted dataset [15]. As a result, we find that these
applied approaches outperform existing state-of-the-art approaches.
|
Styling with Attention to Details
|
Fashion as characterized by its nature, is driven by style. In this paper, we
propose a method that takes into account the style information to complete a
given set of selected fashion items with a complementary fashion item.
Complementary items are those items that can be worn along with the selected
items according to the style. Addressing this problem facilitates in
automatically generating stylish fashion ensembles leading to a richer shopping
experience for users.
Recently, there has been a surge of online social websites where fashion
enthusiasts post the outfit of the day and other users can like and comment on
them. These posts contain a gold-mine of information about style. In this
paper, we exploit these posts to train a deep neural network which captures
style in an automated manner. We pose the problem of predicting complementary
fashion items as a sequence to sequence problem where the input is the selected
set of fashion items and the output is a complementary fashion item based on
the style information learned by the model. We use the encoder decoder
architecture to solve this problem of completing the set of fashion items. We
evaluate the goodness of the proposed model through a variety of experiments.
We empirically observe that our proposed model outperforms competitive baseline
like apriori algorithm by ~28 in terms of accuracy for top-1 recommendation to
complete the fashion ensemble. We also perform retrieval based experiments to
understand the ability of the model to learn style and rank the complementary
fashion items and find that using attention in our encoder decoder model helps
in improving the mean reciprocal rank by ~24. Qualitatively we find the
complementary fashion items generated by our proposed model are richer than the
apriori algorithm.
|
Quantified Markov Logic Networks
|
Markov Logic Networks (MLNs) are well-suited for expressing statistics such
as "with high probability a smoker knows another smoker" but not for expressing
statements such as "there is a smoker who knows most other smokers", which is
necessary for modeling, e.g. influencers in social networks. To overcome this
shortcoming, we study quantified MLNs which generalize MLNs by introducing
statistical universal quantifiers, allowing to express also the latter type of
statistics in a principled way. Our main technical contribution is to show that
the standard reasoning tasks in quantified MLNs, maximum a posteriori and
marginal inference, can be reduced to their respective MLN counterparts in
polynomial time.
|
On decision regions of narrow deep neural networks
|
We show that for neural network functions that have width less or equal to
the input dimension all connected components of decision regions are unbounded.
The result holds for continuous and strictly monotonic activation functions as
well as for the ReLU activation function. This complements recent results on
approximation capabilities by [Hanin 2017 Approximating] and connectivity of
decision regions by [Nguyen 2018 Neural] for such narrow neural networks. Our
results are illustrated by means of numerical experiments.
|
Generating Multi-Categorical Samples with Generative Adversarial
Networks
|
We propose a method to train generative adversarial networks on mutivariate
feature vectors representing multiple categorical values. In contrast to the
continuous domain, where GAN-based methods have delivered considerable results,
GANs struggle to perform equally well on discrete data. We propose and compare
several architectures based on multiple (Gumbel) softmax output layers taking
into account the structure of the data. We evaluate the performance of our
architecture on datasets with different sparsity, number of features, ranges of
categorical values, and dependencies among the features. Our proposed
architecture and method outperforms existing models.
|
Training behavior of deep neural network in frequency domain
|
Why deep neural networks (DNNs) capable of overfitting often generalize well
in practice is a mystery [#zhang2016understanding]. To find a potential
mechanism, we focus on the study of implicit biases underlying the training
process of DNNs. In this work, for both real and synthetic datasets, we
empirically find that a DNN with common settings first quickly captures the
dominant low-frequency components, and then relatively slowly captures the
high-frequency ones. We call this phenomenon Frequency Principle (F-Principle).
The F-Principle can be observed over DNNs of various structures, activation
functions, and training algorithms in our experiments. We also illustrate how
the F-Principle help understand the effect of early-stopping as well as the
generalization of DNNs. This F-Principle potentially provides insights into a
general principle underlying DNN optimization and generalization.
|
Dynamic Control of Explore/Exploit Trade-Off In Bayesian Optimization
|
Bayesian optimization offers the possibility of optimizing black-box
operations not accessible through traditional techniques. The success of
Bayesian optimization methods such as Expected Improvement (EI) are
significantly affected by the degree of trade-off between exploration and
exploitation. Too much exploration can lead to inefficient optimization
protocols, whilst too much exploitation leaves the protocol open to strong
initial biases, and a high chance of getting stuck in a local minimum.
Typically, a constant margin is used to control this trade-off, which results
in yet another hyper-parameter to be optimized. We propose contextual
improvement as a simple, yet effective heuristic to counter this - achieving a
one-shot optimization strategy. Our proposed heuristic can be swiftly
calculated and improves both the speed and robustness of discovery of optimal
solutions. We demonstrate its effectiveness on both synthetic and real world
problems and explore the unaccounted for uncertainty in the pre-determination
of search hyperparameters controlling explore-exploit trade-off.
|
On the Computational Power of Online Gradient Descent
|
We prove that the evolution of weight vectors in online gradient descent can
encode arbitrary polynomial-space computations, even in very simple learning
settings. Our results imply that, under weak complexity-theoretic assumptions,
it is impossible to reason efficiently about the fine-grained behavior of
online gradient descent.
|
Human-level performance in first-person multiplayer games with
population-based deep reinforcement learning
|
Recent progress in artificial intelligence through reinforcement learning
(RL) has shown great success on increasingly complex single-agent environments
and two-player turn-based games. However, the real-world contains multiple
agents, each learning and acting independently to cooperate and compete with
other agents, and environments reflecting this degree of complexity remain an
open challenge. In this work, we demonstrate for the first time that an agent
can achieve human-level in a popular 3D multiplayer first-person video game,
Quake III Arena Capture the Flag, using only pixels and game points as input.
These results were achieved by a novel two-tier optimisation process in which a
population of independent RL agents are trained concurrently from thousands of
parallel matches with agents playing in teams together and against each other
on randomly generated environments. Each agent in the population learns its own
internal reward signal to complement the sparse delayed reward from winning,
and selects actions using a novel temporally hierarchical representation that
enables the agent to reason at multiple timescales. During game-play, these
agents display human-like behaviours such as navigating, following, and
defending based on a rich learned representation that is shown to encode
high-level game knowledge. In an extensive tournament-style evaluation the
trained agents exceeded the win-rate of strong human players both as teammates
and opponents, and proved far stronger than existing state-of-the-art agents.
These results demonstrate a significant jump in the capabilities of artificial
agents, bringing us closer to the goal of human-level intelligence.
|
New Losses for Generative Adversarial Learning
|
Generative Adversarial Networks (Goodfellow et al., 2014), a major
breakthrough in the field of generative modeling, learn a discriminator to
estimate some distance between the target and the candidate distributions.
This paper examines mathematical issues regarding the way the gradients for
the generative model are computed in this context, and notably how to take into
account how the discriminator itself depends on the generator parameters.
A unifying methodology is presented to define mathematically sound training
objectives for generative models taking this dependency into account in a
robust way, covering both GAN, VAE and some GAN variants as particular cases.
|
End-to-End Learning on 3D Protein Structure for Interface Prediction
|
Despite an explosion in the number of experimentally determined, atomically
detailed structures of biomolecules, many critical tasks in structural biology
remain data-limited. Whether performance in such tasks can be improved by using
large repositories of tangentially related structural data remains an open
question. To address this question, we focused on a central problem in biology:
predicting how proteins interact with one another---that is, which surfaces of
one protein bind to those of another protein. We built a training dataset, the
Database of Interacting Protein Structures (DIPS), that contains biases but is
two orders of magnitude larger than those used previously. We found that these
biases significantly degrade the performance of existing methods on
gold-standard data. Hypothesizing that assumptions baked into the hand-crafted
features on which these methods depend were the source of the problem, we
developed the first end-to-end learning model for protein interface prediction,
the Siamese Atomic Surfacelet Network (SASNet). Using only spatial coordinates
and identities of atoms, SASNet outperforms state-of-the-art methods trained on
gold-standard structural data, even when trained on only 3% of our new dataset.
Code and data available at https://github.com/drorlab/DIPS.
|
Generalized Bilinear Deep Convolutional Neural Networks for Multimodal
Biometric Identification
|
In this paper, we propose to employ a bank of modality-dedicated
Convolutional Neural Networks (CNNs), fuse, train, and optimize them together
for person classification tasks. A modality-dedicated CNN is used for each
modality to extract modality-specific features. We demonstrate that, rather
than spatial fusion at the convolutional layers, the fusion can be performed on
the outputs of the fully-connected layers of the modality-specific CNNs without
any loss of performance and with significant reduction in the number of
parameters. We show that, using multiple CNNs with multimodal fusion at the
feature-level, we significantly outperform systems that use unimodal
representation. We study weighted feature, bilinear, and compact bilinear
feature-level fusion algorithms for multimodal biometric person identification.
Finally, We propose generalized compact bilinear fusion algorithm to deploy
both the weighted feature fusion and compact bilinear schemes. We provide the
results for the proposed algorithms on three challenging databases: CMU
Multi-PIE, BioCop, and BIOMDATA.
|
Proceedings of the 2018 ICML Workshop on Human Interpretability in
Machine Learning (WHI 2018)
|
This is the Proceedings of the 2018 ICML Workshop on Human Interpretability
in Machine Learning (WHI 2018), which was held in Stockholm, Sweden, July 14,
2018. Invited speakers were Barbara Engelhardt, Cynthia Rudin, Fernanda
Vi\'egas, and Martin Wattenberg.
|
Multi-Level Feature Abstraction from Convolutional Neural Networks for
Multimodal Biometric Identification
|
In this paper, we propose a deep multimodal fusion network to fuse multiple
modalities (face, iris, and fingerprint) for person identification. The
proposed deep multimodal fusion algorithm consists of multiple streams of
modality-specific Convolutional Neural Networks (CNNs), which are jointly
optimized at multiple feature abstraction levels. Multiple features are
extracted at several different convolutional layers from each modality-specific
CNN for joint feature fusion, optimization, and classification. Features
extracted at different convolutional layers of a modality-specific CNN
represent the input at several different levels of abstract representations. We
demonstrate that an efficient multimodal classification can be accomplished
with a significant reduction in the number of network parameters by exploiting
these multi-level abstract representations extracted from all the
modality-specific CNNs. We demonstrate an increase in multimodal person
identification performance by utilizing the proposed multi-level feature
abstract representations in our multimodal fusion, rather than using only the
features from the last layer of each modality-specific CNNs. We show that our
deep multi-modal CNNs with multimodal fusion at several different feature level
abstraction can significantly outperform the unimodal representation accuracy.
We also demonstrate that the joint optimization of all the modality-specific
CNNs excels the score and decision level fusions of independently optimized
CNNs.
|
Breast Cancer Diagnosis via Classification Algorithms
|
In this paper, we analyze the Wisconsin Diagnostic Breast Cancer Data using
Machine Learning classification techniques, such as the SVM, Bayesian Logistic
Regression (Variational Approximation), and K-Nearest-Neighbors. We describe
each model, and compare their performance through different measures. We
conclude that SVM has the best performance among all other classifiers, while
it competes closely with the Bayesian Logistic Regression that is ranked second
best method for this dataset.
|
COTA: Improving the Speed and Accuracy of Customer Support through
Ranking and Deep Networks
|
For a company looking to provide delightful user experiences, it is of
paramount importance to take care of any customer issues. This paper proposes
COTA, a system to improve speed and reliability of customer support for end
users through automated ticket classification and answers selection for support
representatives. Two machine learning and natural language processing
techniques are demonstrated: one relying on feature engineering (COTA v1) and
the other exploiting raw signals through deep learning architectures (COTA v2).
COTA v1 employs a new approach that converts the multi-classification task into
a ranking problem, demonstrating significantly better performance in the case
of thousands of classes. For COTA v2, we propose an Encoder-Combiner-Decoder, a
novel deep learning architecture that allows for heterogeneous input and output
feature types and injection of prior knowledge through network architecture
choices. This paper compares these models and their variants on the task of
ticket classification and answer selection, showing model COTA v2 outperforms
COTA v1, and analyzes their inner workings and shortcomings. Finally, an A/B
test is conducted in a production setting validating the real-world impact of
COTA in reducing issue resolution time by 10 percent without reducing customer
satisfaction.
|
Anomaly Detection for Skin Disease Images Using Variational Autoencoder
|
In this paper, we demonstrate the potential of applying Variational
Autoencoder (VAE) [10] for anomaly detection in skin disease images. VAE is a
class of deep generative models which is trained by maximizing the evidence
lower bound of data distribution [10]. When trained on only normal data, the
resulting model is able to perform efficient inference and to determine if a
test image is normal or not. We perform experiments on ISIC2018 Challenge
Disease Classification dataset (Task 3) and compare different methods to use
VAE to detect anomaly. The model is able to detect all diseases with 0.779
AUCROC. If we focus on specific diseases, the model is able to detect melanoma
with 0.864 AUCROC and detect actinic keratosis with 0.872 AUCROC, even if it
only sees the images of nevus. To the best of our knowledge, this is the first
applied work of deep generative models for anomaly detection in dermatology.
|
OCTen: Online Compression-based Tensor Decomposition
|
Tensor decompositions are powerful tools for large data analytics as they
jointly model multiple aspects of data into one framework and enable the
discovery of the latent structures and higher-order correlations within the
data. One of the most widely studied and used decompositions, especially in
data mining and machine learning, is the Canonical Polyadic or CP
decomposition. However, today's datasets are not static and these datasets
often dynamically growing and changing with time. To operate on such large
data, we present OCTen the first ever compression-based online parallel
implementation for the CP decomposition. We conduct an extensive empirical
analysis of the algorithms in terms of fitness, memory used and CPU time, and
in order to demonstrate the compression and scalability of the method, we apply
OCTen to big tensor data. Indicatively, OCTen performs on-par or better than
state-of-the-art online and online methods in terms of decomposition accuracy
and efficiency, while saving up to 40-200 % memory space.
|
Efficient ConvNets for Analog Arrays
|
Analog arrays are a promising upcoming hardware technology with the potential
to drastically speed up deep learning. Their main advantage is that they
compute matrix-vector products in constant time, irrespective of the size of
the matrix. However, early convolution layers in ConvNets map very unfavorably
onto analog arrays, because kernel matrices are typically small and the
constant time operation needs to be sequentially iterated a large number of
times, reducing the speed up advantage for ConvNets. Here, we propose to
replicate the kernel matrix of a convolution layer on distinct analog arrays,
and randomly divide parts of the compute among them, so that multiple kernel
matrices are trained in parallel. With this modification, analog arrays execute
ConvNets with an acceleration factor that is proportional to the number of
kernel matrices used per layer (here tested 16-128). Despite having more free
parameters, we show analytically and in numerical experiments that this
convolution architecture is self-regularizing and implicitly learns similar
filters across arrays. We also report superior performance on a number of
datasets and increased robustness to adversarial attacks. Our investigation
suggests to revise the notion that mixed analog-digital hardware is not
suitable for ConvNets.
|
EmbNum: Semantic labeling for numerical values with deep metric learning
|
Semantic labeling for numerical values is a task of assigning semantic labels
to unknown numerical attributes. The semantic labels could be numerical
properties in ontologies, instances in knowledge bases, or labeled data that
are manually annotated by domain experts. In this paper, we refer to semantic
labeling as a retrieval setting where the label of an unknown attribute is
assigned by the label of the most relevant attribute in labeled data. One of
the greatest challenges is that an unknown attribute rarely has the same set of
values with the similar one in the labeled data. To overcome the issue,
statistical interpretation of value distribution is taken into account.
However, the existing studies assume a specific form of distribution. It is not
appropriate in particular to apply open data where there is no knowledge of
data in advance. To address these problems, we propose a neural numerical
embedding model (EmbNum) to learn useful representation vectors for numerical
attributes without prior assumptions on the distribution of data. Then, the
"semantic similarities" between the attributes are measured on these
representation vectors by the Euclidean distance. Our empirical experiments on
City Data and Open Data show that EmbNum significantly outperforms
state-of-the-art methods for the task of numerical attribute semantic labeling
regarding effectiveness and efficiency.
|
Patient representation learning and interpretable evaluation using
clinical notes
|
We have three contributions in this work: 1. We explore the utility of a
stacked denoising autoencoder and a paragraph vector model to learn
task-independent dense patient representations directly from clinical notes. To
analyze if these representations are transferable across tasks, we evaluate
them in multiple supervised setups to predict patient mortality, primary
diagnostic and procedural category, and gender. We compare their performance
with sparse representations obtained from a bag-of-words model. We observe that
the learned generalized representations significantly outperform the sparse
representations when we have few positive instances to learn from, and there is
an absence of strong lexical features. 2. We compare the model performance of
the feature set constructed from a bag of words to that obtained from medical
concepts. In the latter case, concepts represent problems, treatments, and
tests. We find that concept identification does not improve the classification
performance. 3. We propose novel techniques to facilitate model
interpretability. To understand and interpret the representations, we explore
the best encoded features within the patient representations obtained from the
autoencoder model. Further, we calculate feature sensitivity across two
networks to identify the most significant input features for different
classification tasks when we use these pretrained representations as the
supervised input. We successfully extract the most influential features for the
pipeline using this technique.
|
Endmember Extraction on the Grassmannian
|
Endmember extraction plays a prominent role in a variety of data analysis
problems as endmembers often correspond to data representing the purest or best
representative of some feature. Identifying endmembers then can be useful for
further identification and classification tasks. In settings with
high-dimensional data, such as hyperspectral imagery, it can be useful to
consider endmembers that are subspaces as they are capable of capturing a wider
range of variations of a signature. The endmember extraction problem in this
setting thus translates to finding the vertices of the convex hull of a set of
points on a Grassmannian. In the presence of noise, it can be less clear
whether a point should be considered a vertex. In this paper, we propose an
algorithm to extract endmembers on a Grassmannian, identify subspaces of
interest that lie near the boundary of a convex hull, and demonstrate the use
of the algorithm on a synthetic example and on the 220 spectral band AVIRIS
Indian Pines hyperspectral image.
|
Connecting Weighted Automata and Recurrent Neural Networks through
Spectral Learning
|
In this paper, we unravel a fundamental connection between weighted finite
automata~(WFAs) and second-order recurrent neural networks~(2-RNNs): in the
case of sequences of discrete symbols, WFAs and 2-RNNs with linear activation
functions are expressively equivalent. Motivated by this result, we build upon
a recent extension of the spectral learning algorithm to vector-valued WFAs and
propose the first provable learning algorithm for linear 2-RNNs defined over
sequences of continuous input vectors. This algorithm relies on estimating low
rank sub-blocks of the so-called Hankel tensor, from which the parameters of a
linear 2-RNN can be provably recovered. The performances of the proposed method
are assessed in a simulation study.
|
Diagonal Discriminant Analysis with Feature Selection for High
Dimensional Data
|
We introduce a new method of performing high dimensional discriminant
analysis, which we call multiDA. We achieve this by constructing a hybrid model
that seamlessly integrates a multiclass diagonal discriminant analysis model
and feature selection components. Our feature selection component naturally
simplifies to weights which are simple functions of likelihood ratio statistics
allowing natural comparisons with traditional hypothesis testing methods. We
provide heuristic arguments suggesting desirable asymptotic properties of our
algorithm with regards to feature selection. We compare our method with several
other approaches, showing marked improvements in regard to prediction accuracy,
interpretability of chosen features, and algorithm run time. We demonstrate
such strengths of our model by showing strong classification performance on
publicly available high dimensional datasets, as well as through multiple
simulation studies. We make an R package available implementing our approach.
|
SGAD: Soft-Guided Adaptively-Dropped Neural Network
|
Deep neural networks (DNNs) have been proven to have many redundancies.
Hence, many efforts have been made to compress DNNs. However, the existing
model compression methods treat all the input samples equally while ignoring
the fact that the difficulties of various input samples being correctly
classified are different. To address this problem, DNNs with adaptive dropping
mechanism are well explored in this work. To inform the DNNs how difficult the
input samples can be classified, a guideline that contains the information of
input samples is introduced to improve the performance. Based on the developed
guideline and adaptive dropping mechanism, an innovative soft-guided
adaptively-dropped (SGAD) neural network is proposed in this paper. Compared
with the 32 layers residual neural networks, the presented SGAD can reduce the
FLOPs by 77% with less than 1% drop in accuracy on CIFAR-10.
|
Modeling Sparse Deviations for Compressed Sensing using Generative
Models
|
In compressed sensing, a small number of linear measurements can be used to
reconstruct an unknown signal. Existing approaches leverage assumptions on the
structure of these signals, such as sparsity or the availability of a
generative model. A domain-specific generative model can provide a stronger
prior and thus allow for recovery with far fewer measurements. However, unlike
sparsity-based approaches, existing methods based on generative models
guarantee exact recovery only over their support, which is typically only a
small subset of the space on which the signals are defined. We propose
Sparse-Gen, a framework that allows for sparse deviations from the support set,
thereby achieving the best of both worlds by using a domain specific prior and
allowing reconstruction over the full space of signals. Theoretically, our
framework provides a new class of signals that can be acquired using compressed
sensing, reducing classic sparse vector recovery to a special case and avoiding
the restrictive support due to a generative model prior. Empirically, we
observe consistent improvements in reconstruction accuracy over competing
approaches, especially in the more practical setting of transfer compressed
sensing where a generative model for a data-rich, source domain aids sensing on
a data-scarce, target domain.
|
Supervised Reinforcement Learning with Recurrent Neural Network for
Dynamic Treatment Recommendation
|
Dynamic treatment recommendation systems based on large-scale electronic
health records (EHRs) become a key to successfully improve practical clinical
outcomes. Prior relevant studies recommend treatments either use supervised
learning (e.g. matching the indicator signal which denotes doctor
prescriptions), or reinforcement learning (e.g. maximizing evaluation signal
which indicates cumulative reward from survival rates). However, none of these
studies have considered to combine the benefits of supervised learning and
reinforcement learning. In this paper, we propose Supervised Reinforcement
Learning with Recurrent Neural Network (SRL-RNN), which fuses them into a
synergistic learning framework. Specifically, SRL-RNN applies an off-policy
actor-critic framework to handle complex relations among multiple medications,
diseases and individual characteristics. The "actor" in the framework is
adjusted by both the indicator signal and evaluation signal to ensure effective
prescription and low mortality. RNN is further utilized to solve the
Partially-Observed Markov Decision Process (POMDP) problem due to the lack of
fully observed states in real world applications. Experiments on the publicly
real-world dataset, i.e., MIMIC-3, illustrate that our model can reduce the
estimated mortality, while providing promising accuracy in matching doctors'
prescriptions.
|
Empirical fixed point bifurcation analysis
|
In a common experimental setting, the behaviour of a noisy dynamical system
is monitored in response to manipulations of one or more control parameters.
Here, we introduce a structured model to describe parametric changes in
qualitative system behaviour via stochastic bifurcation analysis. In
particular, we describe an extension of Gaussian Process models of transition
maps, in which the learned map is directly parametrized by its fixed points and
associated local linearisations. We show that the system recovers the behaviour
of a well-studied one dimensional system from little data, then learn the
behaviour of a more realistic two dimensional process of mutually inhibiting
neural populations.
|
Factored Bandits
|
We introduce the factored bandits model, which is a framework for learning
with limited (bandit) feedback, where actions can be decomposed into a
Cartesian product of atomic actions. Factored bandits incorporate rank-1
bandits as a special case, but significantly relax the assumptions on the form
of the reward function. We provide an anytime algorithm for stochastic factored
bandits and up to constants matching upper and lower regret bounds for the
problem. Furthermore, we show that with a slight modification the proposed
algorithm can be applied to utility based dueling bandits. We obtain an
improvement in the additive terms of the regret bound compared to state of the
art algorithms (the additive terms are dominating up to time horizons which are
exponential in the number of arms).
|
Generating Synthetic but Plausible Healthcare Record Datasets
|
Generating datasets that "look like" given real ones is an interesting tasks
for healthcare applications of ML and many other fields of science and
engineering. In this paper we propose a new method of general application to
binary datasets based on a method for learning the parameters of a latent
variable moment that we have previously used for clustering patient datasets.
We compare our method with a recent proposal (MedGan) based on generative
adversarial methods and find that the synthetic datasets we generate are
globally more realistic in at least two senses: real and synthetic instances
are harder to tell apart by Random Forests, and the MMD statistic. The most
likely explanation is that our method does not suffer from the "mode collapse"
which is an admitted problem of GANs. Additionally, the generative models we
generate are easy to interpret, unlike the rather obscure GANs. Our experiments
are performed on two patient datasets containing ICD-9 diagnostic codes: the
publicly available MIMIC-III dataset and a dataset containing admissions for
congestive heart failure during 7 years at Hospital de Sant Pau in Barcelona.
|
Curiosity Driven Exploration of Learned Disentangled Goal Spaces
|
Intrinsically motivated goal exploration processes enable agents to
autonomously sample goals to explore efficiently complex environments with
high-dimensional continuous actions. They have been applied successfully to
real world robots to discover repertoires of policies producing a wide
diversity of effects. Often these algorithms relied on engineered goal spaces
but it was recently shown that one can use deep representation learning
algorithms to learn an adequate goal space in simple environments. However, in
the case of more complex environments containing multiple objects or
distractors, an efficient exploration requires that the structure of the goal
space reflects the one of the environment. In this paper we show that using a
disentangled goal space leads to better exploration performances than an
entangled goal space. We further show that when the representation is
disentangled, one can leverage it by sampling goals that maximize learning
progress in a modular manner. Finally, we show that the measure of learning
progress, used to drive curiosity-driven exploration, can be used
simultaneously to discover abstract independently controllable features of the
environment.
|
BIN-CT: Urban Waste Collection based in Predicting the Container Fill
Level
|
The fast demographic growth, together with the concentration of the
population in cities and the increasing amount of daily waste, are factors that
push to the limit the ability of waste assimilation by Nature. Therefore, we
need technological means to make an optimal management of the waste collection
process, which represents 70% of the operational cost in waste treatment. In
this article, we present a free intelligent software system, based on
computational learning algorithms, which plans the best routes for waste
collection supported by past (historical) and future (predictions) data.
The objective of the system is the cost reduction of the waste collection
service by means of the minimization in distance traveled by any truck to
collect a container, hence the fuel consumption. At the same time the quality
of service to the citizen is increased avoiding the annoying overflows of
containers thanks to the accurate fill level predictions performed by BIN-CT.
In this article we show the features of our software system, illustrating it
operation with a real case study of a Spanish city. We conclude that the use of
BIN-CT avoids unnecessary visits to containers, reduces the distance traveled
to collect a container and therefore we obtain a reduction of total costs and
harmful emissions thrown to the atmosphere.
|
Quasi-Monte Carlo Variational Inference
|
Many machine learning problems involve Monte Carlo gradient estimators. As a
prominent example, we focus on Monte Carlo variational inference (MCVI) in this
paper. The performance of MCVI crucially depends on the variance of its
stochastic gradients. We propose variance reduction by means of Quasi-Monte
Carlo (QMC) sampling. QMC replaces N i.i.d. samples from a uniform probability
distribution by a deterministic sequence of samples of length N. This sequence
covers the underlying random variable space more evenly than i.i.d. draws,
reducing the variance of the gradient estimator. With our novel approach, both
the score function and the reparameterization gradient estimators lead to much
faster convergence. We also propose a new algorithm for Monte Carlo objectives,
where we operate with a constant learning rate and increase the number of QMC
samples per iteration. We prove that this way, our algorithm can converge
asymptotically at a faster rate than SGD. We furthermore provide theoretical
guarantees on QMC for Monte Carlo objectives that go beyond MCVI, and support
our findings by several experiments on large-scale data sets from various
domains.
|
Conditional Neural Processes
|
Deep neural networks excel at function approximation, yet they are typically
trained from scratch for each new function. On the other hand, Bayesian
methods, such as Gaussian Processes (GPs), exploit prior knowledge to quickly
infer the shape of a new function at test time. Yet GPs are computationally
expensive, and it can be hard to design appropriate priors. In this paper we
propose a family of neural models, Conditional Neural Processes (CNPs), that
combine the benefits of both. CNPs are inspired by the flexibility of
stochastic processes such as GPs, but are structured as neural networks and
trained via gradient descent. CNPs make accurate predictions after observing
only a handful of training data points, yet scale to complex functions and
large datasets. We demonstrate the performance and versatility of the approach
on a range of canonical machine learning tasks, including regression,
classification and image completion.
|
Ensemble learning with Conformal Predictors: Targeting credible
predictions of conversion from Mild Cognitive Impairment to Alzheimer's
Disease
|
Most machine learning classifiers give predictions for new examples
accurately, yet without indicating how trustworthy predictions are. In the
medical domain, this hampers their integration in decision support systems,
which could be useful in the clinical practice. We use a supervised learning
approach that combines Ensemble learning with Conformal Predictors to predict
conversion from Mild Cognitive Impairment to Alzheimer's Disease. Our goal is
to enhance the classification performance (Ensemble learning) and complement
each prediction with a measure of credibility (Conformal Predictors). Our
results showed the superiority of the proposed approach over a similar ensemble
framework with standard classifiers.
|
Neural Processes
|
A neural network (NN) is a parameterised function that can be tuned via
gradient descent to approximate a labelled collection of data with high
precision. A Gaussian process (GP), on the other hand, is a probabilistic model
that defines a distribution over possible functions, and is updated in light of
data via the rules of probabilistic inference. GPs are probabilistic,
data-efficient and flexible, however they are also computationally intensive
and thus limited in their applicability. We introduce a class of neural latent
variable models which we call Neural Processes (NPs), combining the best of
both worlds. Like GPs, NPs define distributions over functions, are capable of
rapid adaptation to new observations, and can estimate the uncertainty in their
predictions. Like NNs, NPs are computationally efficient during training and
evaluation but also learn to adapt their priors to data. We demonstrate the
performance of NPs on a range of learning tasks, including regression and
optimisation, and compare and contrast with related models in the literature.
|
Privacy Amplification by Subsampling: Tight Analyses via Couplings and
Divergences
|
Differential privacy comes equipped with multiple analytical tools for the
design of private data analyses. One important tool is the so-called "privacy
amplification by subsampling" principle, which ensures that a differentially
private mechanism run on a random subsample of a population provides higher
privacy guarantees than when run on the entire population. Several instances of
this principle have been studied for different random subsampling methods, each
with an ad-hoc analysis. In this paper we present a general method that
recovers and improves prior analyses, yields lower bounds and derives new
instances of privacy amplification by subsampling. Our method leverages a
characterization of differential privacy as a divergence which emerged in the
program verification community. Furthermore, it introduces new tools, including
advanced joint convexity and privacy profiles, which might be of independent
interest.
|
MIXGAN: Learning Concepts from Different Domains for Mixture Generation
|
In this work, we present an interesting attempt on mixture generation:
absorbing different image concepts (e.g., content and style) from different
domains and thus generating a new domain with learned concepts. In particular,
we propose a mixture generative adversarial network (MIXGAN). MIXGAN learns
concepts of content and style from two domains respectively, and thus can join
them for mixture generation in a new domain, i.e., generating images with
content from one domain and style from another. MIXGAN overcomes the limitation
of current GAN-based models which either generate new images in the same domain
as they observed in training stage, or require off-the-shelf content templates
for transferring or translation. Extensive experimental results demonstrate the
effectiveness of MIXGAN as compared to related state-of-the-art GAN-based
models.
|
Encoding Spatial Relations from Natural Language
|
Natural language processing has made significant inroads into learning the
semantics of words through distributional approaches, however representations
learnt via these methods fail to capture certain kinds of information implicit
in the real world. In particular, spatial relations are encoded in a way that
is inconsistent with human spatial reasoning and lacking invariance to
viewpoint changes. We present a system capable of capturing the semantics of
spatial relations such as behind, left of, etc from natural language. Our key
contributions are a novel multi-modal objective based on generating images of
scenes from their textual descriptions, and a new dataset on which to train it.
We demonstrate that internal representations are robust to meaning preserving
transformations of descriptions (paraphrase invariance), while viewpoint
invariance is an emergent property of the system.
|
Ranked Reward: Enabling Self-Play Reinforcement Learning for
Combinatorial Optimization
|
Adversarial self-play in two-player games has delivered impressive results
when used with reinforcement learning algorithms that combine deep neural
networks and tree search. Algorithms like AlphaZero and Expert Iteration learn
tabula-rasa, producing highly informative training data on the fly. However,
the self-play training strategy is not directly applicable to single-player
games. Recently, several practically important combinatorial optimisation
problems, such as the travelling salesman problem and the bin packing problem,
have been reformulated as reinforcement learning problems, increasing the
importance of enabling the benefits of self-play beyond two-player games. We
present the Ranked Reward (R2) algorithm which accomplishes this by ranking the
rewards obtained by a single agent over multiple games to create a relative
performance metric. Results from applying the R2 algorithm to instances of a
two-dimensional and three-dimensional bin packing problems show that it
outperforms generic Monte Carlo tree search, heuristic algorithms and integer
programming solvers. We also present an analysis of the ranked reward
mechanism, in particular, the effects of problem instances with varying
difficulty and different ranking thresholds.
|
Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value
Expansion
|
Integrating model-free and model-based approaches in reinforcement learning
has the potential to achieve the high performance of model-free algorithms with
low sample complexity. However, this is difficult because an imperfect dynamics
model can degrade the performance of the learning algorithm, and in
sufficiently complex environments, the dynamics model will almost always be
imperfect. As a result, a key challenge is to combine model-based approaches
with model-free learning in such a way that errors in the model do not degrade
performance. We propose stochastic ensemble value expansion (STEVE), a novel
model-based technique that addresses this issue. By dynamically interpolating
between model rollouts of various horizon lengths for each individual example,
STEVE ensures that the model is only utilized when doing so does not introduce
significant errors. Our approach outperforms model-free baselines on
challenging continuous control benchmarks with an order-of-magnitude increase
in sample efficiency, and in contrast to previous model-based approaches,
performance does not degrade in complex environments.
|
SPIDER: Near-Optimal Non-Convex Optimization via Stochastic Path
Integrated Differential Estimator
|
In this paper, we propose a new technique named \textit{Stochastic
Path-Integrated Differential EstimatoR} (SPIDER), which can be used to track
many deterministic quantities of interest with significantly reduced
computational cost. We apply SPIDER to two tasks, namely the stochastic
first-order and zeroth-order methods. For stochastic first-order method,
combining SPIDER with normalized gradient descent, we propose two new
algorithms, namely SPIDER-SFO and SPIDER-SFO\textsuperscript{+}, that solve
non-convex stochastic optimization problems using stochastic gradients only. We
provide sharp error-bound results on their convergence rates. In special, we
prove that the SPIDER-SFO and SPIDER-SFO\textsuperscript{+} algorithms achieve
a record-breaking gradient computation cost of $\mathcal{O}\left( \min( n^{1/2}
\epsilon^{-2}, \epsilon^{-3} ) \right)$ for finding an $\epsilon$-approximate
first-order and $\tilde{\mathcal{O}}\left( \min( n^{1/2}
\epsilon^{-2}+\epsilon^{-2.5}, \epsilon^{-3} ) \right)$ for finding an
$(\epsilon, \mathcal{O}(\epsilon^{0.5}))$-approximate second-order stationary
point, respectively. In addition, we prove that SPIDER-SFO nearly matches the
algorithmic lower bound for finding approximate first-order stationary points
under the gradient Lipschitz assumption in the finite-sum setting. For
stochastic zeroth-order method, we prove a cost of $\mathcal{O}( d \min(
n^{1/2} \epsilon^{-2}, \epsilon^{-3}) )$ which outperforms all existing
results.
|
Benchmarking Neural Network Robustness to Common Corruptions and Surface
Variations
|
In this paper we establish rigorous benchmarks for image classifier
robustness. Our first benchmark, ImageNet-C, standardizes and expands the
corruption robustness topic, while showing which classifiers are preferable in
safety-critical applications. Unlike recent robustness research, this benchmark
evaluates performance on commonplace corruptions not worst-case adversarial
corruptions. We find that there are negligible changes in relative corruption
robustness from AlexNet to ResNet classifiers, and we discover ways to enhance
corruption robustness. Then we propose a new dataset called Icons-50 which
opens research on a new kind of robustness, surface variation robustness. With
this dataset we evaluate the frailty of classifiers on new styles of known
objects and unexpected instances of known classes. We also demonstrate two
methods that improve surface variation robustness. Together our benchmarks may
aid future work toward networks that learn fundamental class structure and also
robustly generalize.
|
Restructuring Batch Normalization to Accelerate CNN Training
|
Batch Normalization (BN) has become a core design block of modern
Convolutional Neural Networks (CNNs). A typical modern CNN has a large number
of BN layers in its lean and deep architecture. BN requires mean and variance
calculations over each mini-batch during training. Therefore, the existing
memory access reduction techniques, such as fusing multiple CONV layers, are
not effective for accelerating BN due to their inability to optimize mini-batch
related calculations during training. To address this increasingly important
problem, we propose to restructure BN layers by first splitting a BN layer into
two sub-layers (fission) and then combining the first sub-layer with its
preceding CONV layer and the second sub-layer with the following activation and
CONV layers (fusion). The proposed solution can significantly reduce
main-memory accesses while training the latest CNN models, and the experiments
on a chip multiprocessor show that the proposed BN restructuring can improve
the performance of DenseNet-121 by 25.7%.
|
Transfer Learning for Clinical Time Series Analysis using Recurrent
Neural Networks
|
Deep neural networks have shown promising results for various clinical
prediction tasks such as diagnosis, mortality prediction, predicting duration
of stay in hospital, etc. However, training deep networks -- such as those
based on Recurrent Neural Networks (RNNs) -- requires large labeled data, high
computational resources, and significant hyperparameter tuning effort. In this
work, we investigate as to what extent can transfer learning address these
issues when using deep RNNs to model multivariate clinical time series. We
consider transferring the knowledge captured in an RNN trained on several
source tasks simultaneously using a large labeled dataset to build the model
for a target task with limited labeled data. An RNN pre-trained on several
tasks provides generic features, which are then used to build simpler linear
models for new target tasks without training task-specific RNNs. For
evaluation, we train a deep RNN to identify several patient phenotypes on time
series from MIMIC-III database, and then use the features extracted using that
RNN to build classifiers for identifying previously unseen phenotypes, and also
for a seemingly unrelated task of in-hospital mortality. We demonstrate that
(i) models trained on features extracted using pre-trained RNN outperform or,
in the worst case, perform as well as task-specific RNNs; (ii) the models using
features from pre-trained models are more robust to the size of labeled data
than task-specific RNNs; and (iii) features extracted using pre-trained RNN are
generic enough and perform better than typical statistical hand-crafted
features.
|
Transfer with Model Features in Reinforcement Learning
|
A key question in Reinforcement Learning is which representation an agent can
learn to efficiently reuse knowledge between different tasks. Recently the
Successor Representation was shown to have empirical benefits for transferring
knowledge between tasks with shared transition dynamics. This paper presents
Model Features: a feature representation that clusters behaviourally equivalent
states and that is equivalent to a Model-Reduction. Further, we present a
Successor Feature model which shows that learning Successor Features is
equivalent to learning a Model-Reduction. A novel optimization objective is
developed and we provide bounds showing that minimizing this objective results
in an increasingly improved approximation of a Model-Reduction. Further, we
provide transfer experiments on randomly generated MDPs which vary in their
transition and reward functions but approximately preserve behavioural
equivalence between states. These results demonstrate that Model Features are
suitable for transfer between tasks with varying transition and reward
functions.
|
Proximal algorithms for large-scale statistical modeling and
sensor/actuator selection
|
Several problems in modeling and control of stochastically-driven dynamical
systems can be cast as regularized semi-definite programs. We examine two such
representative problems and show that they can be formulated in a similar
manner. The first, in statistical modeling, seeks to reconcile observed
statistics by suitably and minimally perturbing prior dynamics. The second
seeks to optimally select a subset of available sensors and actuators for
control purposes. To address modeling and control of large-scale systems we
develop a unified algorithmic framework using proximal methods. Our customized
algorithms exploit problem structure and allow handling statistical modeling,
as well as sensor and actuator selection, for substantially larger scales than
what is amenable to current general-purpose solvers. We establish linear
convergence of the proximal gradient algorithm, draw contrast between the
proposed proximal algorithms and alternating direction method of multipliers,
and provide examples that illustrate the merits and effectiveness of our
framework.
|
Understanding and Accelerating Particle-Based Variational Inference
|
Particle-based variational inference methods (ParVIs) have gained attention
in the Bayesian inference literature, for their capacity to yield flexible and
accurate approximations. We explore ParVIs from the perspective of Wasserstein
gradient flows, and make both theoretical and practical contributions. We unify
various finite-particle approximations that existing ParVIs use, and recognize
that the approximation is essentially a compulsory smoothing treatment, in
either of two equivalent forms. This novel understanding reveals the
assumptions and relations of existing ParVIs, and also inspires new ParVIs. We
propose an acceleration framework and a principled bandwidth-selection method
for general ParVIs; these are based on the developed theory and leverage the
geometry of the Wasserstein space. Experimental results show the improved
convergence by the acceleration framework and enhanced sample accuracy by the
bandwidth-selection method.
|
Learning Personalized Representation for Inverse Problems in Medical
Imaging Using Deep Neural Network
|
Recently deep neural networks have been widely and successfully applied in
computer vision tasks and attracted growing interests in medical imaging. One
barrier for the application of deep neural networks to medical imaging is the
need of large amounts of prior training pairs, which is not always feasible in
clinical practice. In this work we propose a personalized representation
learning framework where no prior training pairs are needed, but only the
patient's own prior images. The representation is expressed using a deep neural
network with the patient's prior images as network input. We then applied this
novel image representation to inverse problems in medical imaging in which the
original inverse problem was formulated as a constraint optimization problem
and solved using the alternating direction method of multipliers (ADMM)
algorithm. Anatomically guided brain positron emission tomography (PET) image
reconstruction and image denoising were employed as examples to demonstrate the
effectiveness of the proposed framework. Quantification results based on
simulation and real datasets show that the proposed personalized representation
framework outperform other widely adopted methods.
|
Direct Uncertainty Prediction for Medical Second Opinions
|
The issue of disagreements amongst human experts is a ubiquitous one in both
machine learning and medicine. In medicine, this often corresponds to doctor
disagreements on a patient diagnosis. In this work, we show that machine
learning models can be trained to give uncertainty scores to data instances
that might result in high expert disagreements. In particular, they can
identify patient cases that would benefit most from a medical second opinion.
Our central methodological finding is that Direct Uncertainty Prediction (DUP),
training a model to predict an uncertainty score directly from the raw patient
features, works better than Uncertainty Via Classification, the two-step
process of training a classifier and postprocessing the output distribution to
give an uncertainty score. We show this both with a theoretical result, and on
extensive evaluations on a large scale medical imaging application.
|
BOHB: Robust and Efficient Hyperparameter Optimization at Scale
|
Modern deep learning methods are very sensitive to many hyperparameters, and,
due to the long training times of state-of-the-art models, vanilla Bayesian
hyperparameter optimization is typically computationally infeasible. On the
other hand, bandit-based configuration evaluation approaches based on random
search lack guidance and do not converge to the best configurations as quickly.
Here, we propose to combine the benefits of both Bayesian optimization and
bandit-based methods, in order to achieve the best of both worlds: strong
anytime performance and fast convergence to optimal configurations. We propose
a new practical state-of-the-art hyperparameter optimization method, which
consistently outperforms both Bayesian optimization and Hyperband on a wide
range of problem types, including high-dimensional toy functions, support
vector machines, feed-forward neural networks, Bayesian neural networks, deep
reinforcement learning, and convolutional neural networks. Our method is robust
and versatile, while at the same time being conceptually simple and easy to
implement.
|
Synthetic contrast enhancement in cardiac CT with Deep Learning
|
In Europe the 20% of the CT scans cover the thoracic region. The acquired
images contain information about the cardiovascular system that often remains
latent due to the lack of contrast in the cardiac area. On the other hand, the
contrast enhanced computed tomography (CECT) represents an imaging technique
that allows to easily assess the cardiac chambers volumes and the contrast
dynamics. With this work we aim to face the problem of extraction and
presentation of these latent information, using a deep learning approach with
convolutional neural networks. Starting from the extraction of relevant
features from the image without contrast medium, we try to re-map them on
features typical of CECT, to synthesize an image characterized by an
attenuation in the cardiac chambers as if a virtually iodine contrast medium
was injected. The purposes are to guarantee an estimation of the left cardiac
chambers volume and to perform an evaluation of the contrast dynamics. Our
approach is based on a deconvolutional network trained on a set of 120 patients
who underwent both CT acquisitions in the same contrastographic arterial phase
and the same cardiac phase. To ensure a reliable predicted CECT image, in terms
of values and morphology, a custom loss function is defined by combining an
error function to find a pixel-wise correspondence, which takes into account
the similarity in term of Hounsfield units between the input and output images
and by a cross-entropy computed on the binarized versions of the synthesized
and of the real CECT image. The proposed method is finally tested on 20
subjects.
|
Program Language Translation Using a Grammar-Driven Tree-to-Tree Model
|
The task of translating between programming languages differs from the
challenge of translating natural languages in that programming languages are
designed with a far more rigid set of structural and grammatical rules.
Previous work has used a tree-to-tree encoder/decoder model to take advantage
of the inherent tree structure of programs during translation. Neural decoders,
however, by default do not exploit known grammar rules of the target language.
In this paper, we describe a tree decoder that leverages knowledge of a
language's grammar rules to exclusively generate syntactically correct
programs. We find that this grammar-based tree-to-tree model outperforms the
state of the art tree-to-tree model in translating between two programming
languages on a previously used synthetic task.
|
Regularizing Autoencoder-Based Matrix Completion Models via Manifold
Learning
|
Autoencoders are popular among neural-network-based matrix completion models
due to their ability to retrieve potential latent factors from the partially
observed matrices. Nevertheless, when training data is scarce their performance
is significantly degraded due to overfitting. In this paper, we mit- igate
overfitting with a data-dependent regularization technique that relies on the
principles of multi-task learning. Specifically, we propose an
autoencoder-based matrix completion model that performs prediction of the
unknown matrix values as a main task, and manifold learning as an auxiliary
task. The latter acts as an inductive bias, leading to solutions that
generalize better. The proposed model outperforms the existing
autoencoder-based models designed for matrix completion, achieving high
reconstruction accuracy in well-known datasets.
|
Discrete Sampling using Semigradient-based Product Mixtures
|
We consider the problem of inference in discrete probabilistic models, that
is, distributions over subsets of a finite ground set. These encompass a range
of well-known models in machine learning, such as determinantal point processes
and Ising models. Locally-moving Markov chain Monte Carlo algorithms, such as
the Gibbs sampler, are commonly used for inference in such models, but their
convergence is, at times, prohibitively slow. This is often caused by
state-space bottlenecks that greatly hinder the movement of such samplers. We
propose a novel sampling strategy that uses a specific mixture of product
distributions to propose global moves and, thus, accelerate convergence.
Furthermore, we show how to construct such a mixture using semigradient
information. We illustrate the effectiveness of combining our sampler with
existing ones, both theoretically on an example model, as well as practically
on three models learned from real-world data sets.
|
Learning Theory and Algorithms for Revenue Management in Sponsored
Search
|
Online advertisement is the main source of revenue for Internet business.
Advertisers are typically ranked according to a score that takes into account
their bids and potential click-through rates(eCTR). Generally, the likelihood
that a user clicks on an ad is often modeled by optimizing for the click
through rates rather than the performance of the auction in which the click
through rates will be used. This paper attempts to eliminate this
dis-connection by proposing loss functions for click modeling that are based on
final auction performance.In this paper, we address two feasible metrics (AUC^R
and SAUC) to evaluate the on-line RPM (revenue per mille) directly rather than
the CTR. And then, we design an explicit ranking function by incorporating the
calibration fac-tor and price-squashed factor to maximize the revenue. Given
the power of deep networks, we also explore an implicit optimal ranking
function with deep model. Lastly, various experiments with two real world
datasets are presented. In particular, our proposed methods perform better than
the state-of-the-art methods with regard to the revenue of the platform.
|
Per-decision Multi-step Temporal Difference Learning with Control
Variates
|
Multi-step temporal difference (TD) learning is an important approach in
reinforcement learning, as it unifies one-step TD learning with Monte Carlo
methods in a way where intermediate algorithms can outperform either extreme.
They address a bias-variance trade off between reliance on current estimates,
which could be poor, and incorporating longer sampled reward sequences into the
updates. Especially in the off-policy setting, where the agent aims to learn
about a policy different from the one generating its behaviour, the variance in
the updates can cause learning to diverge as the number of sampled rewards used
in the estimates increases. In this paper, we introduce per-decision control
variates for multi-step TD algorithms, and compare them to existing methods.
Our results show that including the control variates can greatly improve
performance on both on and off-policy multi-step temporal difference learning
tasks.
|
Logistic Regression, Neural Networks and Dempster-Shafer Theory: a New
Perspective
|
We revisit logistic regression and its nonlinear extensions, including
multilayer feedforward neural networks, by showing that these classifiers can
be viewed as converting input or higher-level features into Dempster-Shafer
mass functions and aggregating them by Dempster's rule of combination. The
probabilistic outputs of these classifiers are the normalized plausibilities
corresponding to the underlying combined mass function. This mass function is
more informative than the output probability distribution. In particular, it
makes it possible to distinguish between lack of evidence (when none of the
features provides discriminant information) from conflicting evidence (when
different features support different classes). This expressivity of mass
functions allows us to gain insight into the role played by each input feature
in logistic regression, and to interpret hidden unit outputs in multilayer
neural networks. It also makes it possible to use alternative decision rules,
such as interval dominance, which select a set of classes when the available
evidence does not unambiguously point to a single class, thus trading reduced
error rate for higher imprecision.
|
Privacy-preserving Machine Learning through Data Obfuscation
|
As machine learning becomes a practice and commodity, numerous cloud-based
services and frameworks are provided to help customers develop and deploy
machine learning applications. While it is prevalent to outsource model
training and serving tasks in the cloud, it is important to protect the privacy
of sensitive samples in the training dataset and prevent information leakage to
untrusted third parties. Past work have shown that a malicious machine learning
service provider or end user can easily extract critical information about the
training samples, from the model parameters or even just model outputs.
In this paper, we propose a novel and generic methodology to preserve the
privacy of training data in machine learning applications. Specifically we
introduce an obfuscate function and apply it to the training data before
feeding them to the model training task. This function adds random noise to
existing samples, or augments the dataset with new samples. By doing so
sensitive information about the properties of individual samples, or
statistical properties of a group of samples, is hidden. Meanwhile the model
trained from the obfuscated dataset can still achieve high accuracy. With this
approach, the customers can safely disclose the data or models to third-party
providers or end users without the need to worry about data privacy. Our
experiments show that this approach can effective defeat four existing types of
machine learning privacy attacks at negligible accuracy cost.
|
Hunting the Ethereum Smart Contract: Color-inspired Inspection of
Potential Attacks
|
Blockchain and Cryptocurrencies are gaining unprecedented popularity and
understanding. Meanwhile, Ethereum is gaining a significant popularity in the
blockchain community, mainly due to the fact that it is designed in a way that
enables developers to write smart contract and decentralized applications
(Dapps). This new paradigm of applications opens the door to many possibilities
and opportunities. However, the security of Ethereum smart contracts has not
received much attention; several Ethereum smart contracts malfunctioning have
recently been reported. Unlike many previous works that have applied static and
dynamic analyses to find bugs in smart contracts, we do not attempt to define
and extract any features; instead we focus on reducing the expert's labor
costs. We first present a new in-depth analysis of potential attacks
methodology and then translate the bytecode of solidity into RGB color code.
After that, we transform them to a fixed-sized encoded image. Finally, the
encoded image is fed to convolutional neural network (CNN) for automatic
feature extraction and learning, detecting compiler bugs of Ethereum smart
contract.
|
Learning in Variational Autoencoders with Kullback-Leibler and Renyi
Integral Bounds
|
In this paper we propose two novel bounds for the log-likelihood based on
Kullback-Leibler and the R\'{e}nyi divergences, which can be used for
variational inference and in particular for the training of Variational
AutoEncoders. Our proposal is motivated by the difficulties encountered in
training VAEs on continuous datasets with high contrast images, such as those
with handwritten digits and characters, where numerical issues often appear
unless noise is added, either to the dataset during training or to the
generative model given by the decoder. The new bounds we propose, which are
obtained from the maximization of the likelihood of an interval for the
observations, allow numerically stable training procedures without the
necessity of adding any extra source of noise to the data.
|
Neural Language Codes for Multilingual Acoustic Models
|
Multilingual Speech Recognition is one of the most costly AI problems,
because each language (7,000+) and even different accents require their own
acoustic models to obtain best recognition performance. Even though they all
use the same phoneme symbols, each language and accent imposes its own coloring
or "twang". Many adaptive approaches have been proposed, but they require
further training, additional data and generally are inferior to monolingually
trained models. In this paper, we propose a different approach that uses a
large multilingual model that is \emph{modulated} by the codes generated by an
ancillary network that learns to code useful differences between the "twangs"
or human language.
We use Meta-Pi networks to have one network (the language code net) gate the
activity of neurons in another (the acoustic model nets). Our results show that
during recognition multilingual Meta-Pi networks quickly adapt to the proper
language coloring without retraining or new data, and perform better than
monolingually trained networks. The model was evaluated by training acoustic
modeling nets and modulating language code nets jointly and optimize them for
best recognition performance.
|
Deeply-Sparse Signal rePresentations ($\text{D}\text{S}^2\text{P}$)
|
A recent line of work shows that a deep neural network with ReLU
nonlinearities arises from a finite sequence of cascaded sparse coding models,
the outputs of which, except for the last element in the cascade, are sparse
and unobservable. That is, intermediate outputs deep in the cascade are sparse,
hence the title of this manuscript. We show here, using techniques from the
dictionary learning literature that, if the measurement matrices in the
cascaded sparse coding model (a) satisfy RIP and (b) all have sparse columns
except for the last, they can be recovered with high probability. We propose
two algorithms for this purpose: one that recovers the matrices in a forward
sequence, and another that recovers them in a backward sequence. The method of
choice in deep learning to solve this problem is by training an auto-encoder.
Our algorithms provide a sound alternative, with theoretical guarantees, as
well upper bounds on sample complexity. The theory shows that the learning
complexity of the forward algorithm depends on the number of hidden units at
the deepest layer and the number of active neurons at that layer (sparsity). In
addition, the theory relates the number of hidden units in successive layers,
thus giving a practical prescription for designing deep ReLU neural networks.
Because it puts fewer restrictions on the architecture, the backward algorithm
requires more data. We demonstrate the deep dictionary learning algorithm via
simulations. Finally, we use a coupon-collection argument to conjecture a lower
bound on sample complexity that gives some insight as to why deep networks
require more data to train than shallow ones.
|
Deep Reinforcement Learning for Doom using Unsupervised Auxiliary Tasks
|
Recent developments in deep reinforcement learning have enabled the creation
of agents for solving a large variety of games given a visual input. These
methods have been proven successful for 2D games, like the Atari games, or for
simple tasks, like navigating in mazes. It is still an open question, how to
address more complex environments, in which the reward is sparse and the state
space is huge. In this paper we propose a divide and conquer deep reinforcement
learning solution and we test our agent in the first person shooter (FPS) game
of Doom. Our work is based on previous works in deep reinforcement learning and
in Doom agents. We also present how our agent is able to perform better in
unknown environments compared to a state of the art reinforcement learning
algorithm.
|
A Boo(n) for Evaluating Architecture Performance
|
We point out important problems with the common practice of using the best
single model performance for comparing deep learning architectures, and we
propose a method that corrects these flaws. Each time a model is trained, one
gets a different result due to random factors in the training process, which
include random parameter initialization and random data shuffling. Reporting
the best single model performance does not appropriately address this
stochasticity. We propose a normalized expected best-out-of-$n$ performance
($\text{Boo}_n$) as a way to correct these problems.
|
Variational Bayesian dropout: pitfalls and fixes
|
Dropout, a stochastic regularisation technique for training of neural
networks, has recently been reinterpreted as a specific type of approximate
inference algorithm for Bayesian neural networks. The main contribution of the
reinterpretation is in providing a theoretical framework useful for analysing
and extending the algorithm. We show that the proposed framework suffers from
several issues; from undefined or pathological behaviour of the true posterior
related to use of improper priors, to an ill-defined variational objective due
to singularity of the approximating distribution relative to the true
posterior. Our analysis of the improper log uniform prior used in variational
Gaussian dropout suggests the pathologies are generally irredeemable, and that
the algorithm still works only because the variational formulation annuls some
of the pathologies. To address the singularity issue, we proffer Quasi-KL (QKL)
divergence, a new approximate inference objective for approximation of
high-dimensional distributions. We show that motivations for variational
Bernoulli dropout based on discretisation and noise have QKL as a limit.
Properties of QKL are studied both theoretically and on a simple practical
example which shows that the QKL-optimal approximation of a full rank Gaussian
with a degenerate one naturally leads to the Principal Component Analysis
solution.
|
Arcades: A deep model for adaptive decision making in voice controlled
smart-home
|
In a voice-controlled smart-home, a controller must respond not only to
user's requests but also according to the interaction context. This paper
describes Arcades, a system which uses deep reinforcement learning to extract
context from a graphical representation of home automation system and to update
continuously its behavior to the user's one. This system is robust to changes
in the environment (sensor breakdown or addition) through its graphical
representation (scale well) and the reinforcement mechanism (adapt well). The
experiments on realistic data demonstrate that this method promises to reach
long life context-aware control of smart-home.
|
Beef Cattle Instance Segmentation Using Fully Convolutional Neural
Network
|
We present an instance segmentation algorithm trained and applied to a CCTV
recording of beef cattle during a winter finishing period. A fully
convolutional network was transformed into an instance segmentation network
that learns to label each instance of an animal separately. We introduce a
conceptually simple framework that the network uses to output a single
prediction for every animal. These results are a contribution towards behaviour
analysis in winter finishing beef cattle for early detection of animal
welfare-related problems.
|
BayesGrad: Explaining Predictions of Graph Convolutional Networks
|
Recent advances in graph convolutional networks have significantly improved
the performance of chemical predictions, raising a new research question: "how
do we explain the predictions of graph convolutional networks?" A possible
approach to answer this question is to visualize evidence substructures
responsible for the predictions. For chemical property prediction tasks, the
sample size of the training data is often small and/or a label imbalance
problem occurs, where a few samples belong to a single class and the majority
of samples belong to the other classes. This can lead to uncertainty related to
the learned parameters of the machine learning model. To address this
uncertainty, we propose BayesGrad, utilizing the Bayesian predictive
distribution, to define the importance of each node in an input graph, which is
computed efficiently using the dropout technique. We demonstrate that BayesGrad
successfully visualizes the substructures responsible for the label prediction
in the artificial experiment, even when the sample size is small. Furthermore,
we use a real dataset to evaluate the effectiveness of the visualization. The
basic idea of BayesGrad is not limited to graph-structured data and can be
applied to other data types.
|
Improving Unsupervised Defect Segmentation by Applying Structural
Similarity to Autoencoders
|
Convolutional autoencoders have emerged as popular methods for unsupervised
defect segmentation on image data. Most commonly, this task is performed by
thresholding a pixel-wise reconstruction error based on an $\ell^p$ distance.
This procedure, however, leads to large residuals whenever the reconstruction
encompasses slight localization inaccuracies around edges. It also fails to
reveal defective regions that have been visually altered when intensity values
stay roughly consistent. We show that these problems prevent these approaches
from being applied to complex real-world scenarios and that it cannot be easily
avoided by employing more elaborate architectures such as variational or
feature matching autoencoders. We propose to use a perceptual loss function
based on structural similarity which examines inter-dependencies between local
image regions, taking into account luminance, contrast and structural
information, instead of simply comparing single pixel values. It achieves
significant performance gains on a challenging real-world dataset of
nanofibrous materials and a novel dataset of two woven fabrics over the state
of the art approaches for unsupervised defect segmentation that use pixel-wise
reconstruction error metrics.
|
Consistent Generative Query Networks
|
Stochastic video prediction models take in a sequence of image frames, and
generate a sequence of consecutive future image frames. These models typically
generate future frames in an autoregressive fashion, which is slow and requires
the input and output frames to be consecutive. We introduce a model that
overcomes these drawbacks by generating a latent representation from an
arbitrary set of frames that can then be used to simultaneously and efficiently
sample temporally consistent frames at arbitrary time-points. For example, our
model can "jump" and directly sample frames at the end of the video, without
sampling intermediate frames. Synthetic video evaluations confirm substantial
gains in speed and functionality without loss in fidelity. We also apply our
framework to a 3D scene reconstruction dataset. Here, our model is conditioned
on camera location and can sample consistent sets of images for what an
occluded region of a 3D scene might look like, even if there are multiple
possibilities for what that region might contain. Reconstructions and videos
are available at https://bit.ly/2O4Pc4R.
|
TFLMS: Large Model Support in TensorFlow by Graph Rewriting
|
While accelerators such as GPUs have limited memory, deep neural networks are
becoming larger and will not fit with the memory limitation of accelerators for
training. We propose an approach to tackle this problem by rewriting the
computational graph of a neural network, in which swap-out and swap-in
operations are inserted to temporarily store intermediate results on CPU
memory. In particular, we first revise the concept of a computational graph by
defining a concrete semantics for variables in a graph. We then formally show
how to derive swap-out and swap-in operations from an existing graph and
present rules to optimize the graph. To realize our approach, we developed a
module in TensorFlow, named TFLMS. TFLMS is published as a pull request in the
TensorFlow repository for contributing to the TensorFlow community. With TFLMS,
we were able to train ResNet-50 and 3DUnet with 4.7x and 2x larger batch size,
respectively. In particular, we were able to train 3DUNet using images of size
of $192^3$ for image segmentation, which, without TFLMS, had been done only by
dividing the images to smaller images, which affects the accuracy.
|
Goal-oriented Trajectories for Efficient Exploration
|
Exploration is a difficult challenge in reinforcement learning and even
recent state-of-the art curiosity-based methods rely on the simple
epsilon-greedy strategy to generate novelty. We argue that pure random walks do
not succeed to properly expand the exploration area in most environments and
propose to replace single random action choices by random goals selection
followed by several steps in their direction. This approach is compatible with
any curiosity-based exploration and off-policy reinforcement learning agents
and generates longer and safer trajectories than individual random actions. To
illustrate this, we present a task-independent agent that learns to reach
coordinates in screen frames and demonstrate its ability to explore with the
game Super Mario Bros. improving significantly the score of a baseline DQN
agent.
|
Linear Bandits with Stochastic Delayed Feedback
|
Stochastic linear bandits are a natural and well-studied model for structured
exploration/exploitation problems and are widely used in applications such as
online marketing and recommendation. One of the main challenges faced by
practitioners hoping to apply existing algorithms is that usually the feedback
is randomly delayed and delays are only partially observable. For example,
while a purchase is usually observable some time after the display, the
decision of not buying is never explicitly sent to the system. In other words,
the learner only observes delayed positive events. We formalize this problem as
a novel stochastic delayed linear bandit and propose ${\tt OTFLinUCB}$ and
${\tt OTFLinTS}$, two computationally efficient algorithms able to integrate
new information as it becomes available and to deal with the permanently
censored feedback. We prove optimal $\tilde O(\smash{d\sqrt{T}})$ bounds on the
regret of the first algorithm and study the dependency on delay-dependent
parameters. Our model, assumptions and results are validated by experiments on
simulated and real data.
|
Scalable Gaussian Processes with Grid-Structured Eigenfunctions
(GP-GRIEF)
|
We introduce a kernel approximation strategy that enables computation of the
Gaussian process log marginal likelihood and all hyperparameter derivatives in
$\mathcal{O}(p)$ time. Our GRIEF kernel consists of $p$ eigenfunctions found
using a Nystrom approximation from a dense Cartesian product grid of inducing
points. By exploiting algebraic properties of Kronecker and Khatri-Rao tensor
products, computational complexity of the training procedure can be practically
independent of the number of inducing points. This allows us to use arbitrarily
many inducing points to achieve a globally accurate kernel approximation, even
in high-dimensional problems. The fast likelihood evaluation enables type-I or
II Bayesian inference on large-scale datasets. We benchmark our algorithms on
real-world problems with up to two-million training points and $10^{33}$
inducing points.
|
Adaptive Path-Integral Autoencoder: Representation Learning and Planning
for Dynamical Systems
|
We present a representation learning algorithm that learns a low-dimensional
latent dynamical system from high-dimensional \textit{sequential} raw data,
e.g., video. The framework builds upon recent advances in amortized inference
methods that use both an inference network and a refinement procedure to output
samples from a variational distribution given an observation sequence, and
takes advantage of the duality between control and inference to approximately
solve the intractable inference problem using the path integral control
approach. The learned dynamical model can be used to predict and plan the
future states; we also present the efficient planning method that exploits the
learned low-dimensional latent dynamics. Numerical experiments show that the
proposed path-integral control based variational inference method leads to
tighter lower bounds in statistical model learning of sequential data. The
supplementary video: https://youtu.be/xCp35crUoLQ
|
Scalable Recommender Systems through Recursive Evidence Chains
|
Recommender systems can be formulated as a matrix completion problem,
predicting ratings from user and item parameter vectors. Optimizing these
parameters by subsampling data becomes difficult as the number of users and
items grows. We develop a novel approach to generate all latent variables on
demand from the ratings matrix itself and a fixed pool of parameters. We
estimate missing ratings using chains of evidence that link them to a small set
of prototypical users and items. Our model automatically addresses the
cold-start and online learning problems by combining information across both
users and items. We investigate the scaling behavior of this model, and
demonstrate competitive results with respect to current matrix factorization
techniques in terms of accuracy and convergence speed.
|
V-CNN: When Convolutional Neural Network encounters Data Visualization
|
In recent years, deep learning poses a deep technical revolution in almost
every field and attracts great attentions from industry and academia.
Especially, the convolutional neural network (CNN), one representative model of
deep learning, achieves great successes in computer vision and natural language
processing. However, simply or blindly applying CNN to the other fields results
in lower training effects or makes it quite difficult to adjust the model
parameters. In this poster, we propose a general methodology named V-CNN by
introducing data visualizing for CNN. V-CNN introduces a data visualization
model prior to CNN modeling to make sure the data after processing is fit for
the features of images as well as CNN modeling. We apply V-CNN to the network
intrusion detection problem based on a famous practical dataset: AWID.
Simulation results confirm V-CNN significantly outperforms other studies and
the recall rate of each invasion category is more than 99.8%.
|
Encoding Motion Primitives for Autonomous Vehicles using Virtual
Velocity Constraints and Neural Network Scheduling
|
Within the context of trajectory planning for autonomous vehicles this paper
proposes methods for efficient encoding of motion primitives in neural networks
on top of model-based and gradient-free reinforcement learning. It is
distinguished between 5 core aspects: system model, network architecture,
training algorithm, training tasks selection and hardware/software
implementation. For the system model, a kinematic (3-states-2-controls) and a
dynamic (16-states-2-controls) vehicle model are compared. For the network
architecture, 3 feedforward structures are compared including weighted skip
connections. For the training algorithm, virtual velocity constraints and
network scheduling are proposed. For the training tasks, different feature
vector selections are discussed. For the implementation, aspects of
gradient-free learning using 1 GPU and the handling of perturbation noise
therefore are discussed. The effects of proposed methods are illustrated in
experiments encoding up to 14625 motion primitives. The capabilities of tiny
neural networks with as few as 10 scalar parameters when scheduled on vehicle
velocity are emphasized.
|
Implicit Generative Modeling of Random Noise during Training for
Adversarial Robustness
|
We introduce a Noise-based prior Learning (NoL) approach for training neural
networks that are intrinsically robust to adversarial attacks. We find that the
implicit generative modeling of random noise with the same loss function used
during posterior maximization, improves a model's understanding of the data
manifold furthering adversarial robustness. We evaluate our approach's efficacy
and provide a simplistic visualization tool for understanding adversarial data,
using Principal Component Analysis. Our analysis reveals that adversarial
robustness, in general, manifests in models with higher variance along the
high-ranked principal components. We show that models learnt with our approach
perform remarkably well against a wide-range of attacks. Furthermore, combining
NoL with state-of-the-art adversarial training extends the robustness of a
model, even beyond what it is adversarially trained for, in both white-box and
black-box attack scenarios.
|
U-SLADS: Unsupervised Learning Approach for Dynamic Dendrite Sampling
|
Novel data acquisition schemes have been an emerging need for scanning
microscopy based imaging techniques to reduce the time in data acquisition and
to minimize probing radiation in sample exposure. Varies sparse sampling
schemes have been studied and are ideally suited for such applications where
the images can be reconstructed from a sparse set of measurements. Dynamic
sparse sampling methods, particularly supervised learning based iterative
sampling algorithms, have shown promising results for sampling pixel locations
on the edges or boundaries during imaging. However, dynamic sampling for
imaging skeleton-like objects such as metal dendrites remains difficult. Here,
we address a new unsupervised learning approach using Hierarchical Gaussian
Mixture Mod- els (HGMM) to dynamically sample metal dendrites. This technique
is very useful if the users are interested in fast imaging the primary and
secondary arms of metal dendrites in solidification process in materials
science.
|
Distributed Self-Paced Learning in Alternating Direction Method of
Multipliers
|
Self-paced learning (SPL) mimics the cognitive process of humans, who
generally learn from easy samples to hard ones. One key issue in SPL is the
training process required for each instance weight depends on the other samples
and thus cannot easily be run in a distributed manner in a large-scale dataset.
In this paper, we reformulate the self-paced learning problem into a
distributed setting and propose a novel Distributed Self-Paced Learning method
(DSPL) to handle large-scale datasets. Specifically, both the model and
instance weights can be optimized in parallel for each batch based on a
consensus alternating direction method of multipliers. We also prove the
convergence of our algorithm under mild conditions. Extensive experiments on
both synthetic and real datasets demonstrate that our approach is superior to
those of existing methods.
|
Towards more Reliable Transfer Learning
|
Multi-source transfer learning has been proven effective when within-target
labeled data is scarce. Previous work focuses primarily on exploiting domain
similarities and assumes that source domains are richly or at least comparably
labeled. While this strong assumption is never true in practice, this paper
relaxes it and addresses challenges related to sources with diverse labeling
volume and diverse reliability. The first challenge is combining domain
similarity and source reliability by proposing a new transfer learning method
that utilizes both source-target similarities and inter-source relationships.
The second challenge involves pool-based active learning where the oracle is
only available in source domains, resulting in an integrated active transfer
learning framework that incorporates distribution matching and uncertainty
sampling. Extensive experiments on synthetic and two real-world datasets
clearly demonstrate the superiority of our proposed methods over several
baselines including state-of-the-art transfer learning methods.
|
Variance Reduction for Reinforcement Learning in Input-Driven
Environments
|
We consider reinforcement learning in input-driven environments, where an
exogenous, stochastic input process affects the dynamics of the system. Input
processes arise in many applications, including queuing systems, robotics
control with disturbances, and object tracking. Since the state dynamics and
rewards depend on the input process, the state alone provides limited
information for the expected future returns. Therefore, policy gradient methods
with standard state-dependent baselines suffer high variance during training.
We derive a bias-free, input-dependent baseline to reduce this variance, and
analytically show its benefits over state-dependent baselines. We then propose
a meta-learning approach to overcome the complexity of learning a baseline that
depends on a long sequence of inputs. Our experimental results show that across
environments from queuing systems, computer networks, and MuJoCo robotic
locomotion, input-dependent baselines consistently improve training stability
and result in better eventual policies.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.