title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Zoom Better to See Clearer: Human and Object Parsing with Hierarchical
Auto-Zoom Net | Parsing articulated objects, e.g. humans and animals, into semantic parts
(e.g. body, head and arms, etc.) from natural images is a challenging and
fundamental problem for computer vision. A big difficulty is the large
variability of scale and location for objects and their corresponding parts.
Even limited mistakes in estimating scale and location will degrade the parsing
output and cause errors in boundary details. To tackle these difficulties, we
propose a "Hierarchical Auto-Zoom Net" (HAZN) for object part parsing which
adapts to the local scales of objects and parts. HAZN is a sequence of two
"Auto-Zoom Net" (AZNs), each employing fully convolutional networks that
perform two tasks: (1) predict the locations and scales of object instances
(the first AZN) or their parts (the second AZN); (2) estimate the part scores
for predicted object instance or part regions. Our model can adaptively "zoom"
(resize) predicted image regions into their proper scales to refine the
parsing.
We conduct extensive experiments over the PASCAL part datasets on humans,
horses, and cows. For humans, our approach significantly outperforms the
state-of-the-arts by 5% mIOU and is especially better at segmenting small
instances and small parts. We obtain similar improvements for parsing cows and
horses over alternative methods. In summary, our strategy of first zooming into
objects and then zooming into parts is very effective. It also enables us to
process different regions of the image at different scales adaptively so that,
for example, we do not need to waste computational resources scaling the entire
image.
|
Gaussian Process Planning with Lipschitz Continuous Reward Functions:
Towards Unifying Bayesian Optimization, Active Learning, and Beyond | This paper presents a novel nonmyopic adaptive Gaussian process planning
(GPP) framework endowed with a general class of Lipschitz continuous reward
functions that can unify some active learning/sensing and Bayesian optimization
criteria and offer practitioners some flexibility to specify their desired
choices for defining new tasks/problems. In particular, it utilizes a
principled Bayesian sequential decision problem framework for jointly and
naturally optimizing the exploration-exploitation trade-off. In general, the
resulting induced GPP policy cannot be derived exactly due to an uncountable
set of candidate observations. A key contribution of our work here thus lies in
exploiting the Lipschitz continuity of the reward functions to solve for a
nonmyopic adaptive epsilon-optimal GPP (epsilon-GPP) policy. To plan in real
time, we further propose an asymptotically optimal, branch-and-bound anytime
variant of epsilon-GPP with performance guarantee. We empirically demonstrate
the effectiveness of our epsilon-GPP policy and its anytime variant in Bayesian
optimization and an energy harvesting task.
|
Near-Optimal Active Learning of Multi-Output Gaussian Processes | This paper addresses the problem of active learning of a multi-output
Gaussian process (MOGP) model representing multiple types of coexisting
correlated environmental phenomena. In contrast to existing works, our active
learning problem involves selecting not just the most informative sampling
locations to be observed but also the types of measurements at each selected
location for minimizing the predictive uncertainty (i.e., posterior joint
entropy) of a target phenomenon of interest given a sampling budget.
Unfortunately, such an entropy criterion scales poorly in the numbers of
candidate sampling locations and selected observations when optimized. To
resolve this issue, we first exploit a structure common to sparse MOGP models
for deriving a novel active learning criterion. Then, we exploit a relaxed form
of submodularity property of our new criterion for devising a polynomial-time
approximation algorithm that guarantees a constant-factor approximation of that
achieved by the optimal set of selected observations. Empirical evaluation on
real-world datasets shows that our proposed approach outperforms existing
algorithms for active learning of MOGP and single-output GP models.
|
BlackOut: Speeding up Recurrent Neural Network Language Models With Very
Large Vocabularies | We propose BlackOut, an approximation algorithm to efficiently train massive
recurrent neural network language models (RNNLMs) with million word
vocabularies. BlackOut is motivated by using a discriminative loss, and we
describe a new sampling strategy which significantly reduces computation while
improving stability, sample efficiency, and rate of convergence. One way to
understand BlackOut is to view it as an extension of the DropOut strategy to
the output layer, wherein we use a discriminative training loss and a weighted
sampling scheme. We also establish close connections between BlackOut,
importance sampling, and noise contrastive estimation (NCE). Our experiments,
on the recently released one billion word language modeling benchmark,
demonstrate scalability and accuracy of BlackOut; we outperform the
state-of-the art, and achieve the lowest perplexity scores on this dataset.
Moreover, unlike other established methods which typically require GPUs or CPU
clusters, we show that a carefully implemented version of BlackOut requires
only 1-10 days on a single machine to train a RNNLM with a million word
vocabulary and billions of parameters on one billion words. Although we
describe BlackOut in the context of RNNLM training, it can be used to any
networks with large softmax output layers.
|
ICU Patient Deterioration prediction: a Data-Mining Approach | A huge amount of medical data is generated every day, which presents a
challenge in analysing these data. The obvious solution to this challenge is to
reduce the amount of data without information loss. Dimension reduction is
considered the most popular approach for reducing data size and also to reduce
noise and redundancies in data. In this paper, we investigate the effect of
feature selection in improving the prediction of patient deterioration in ICUs.
We consider lab tests as features. Thus, choosing a subset of features would
mean choosing the most important lab tests to perform. If the number of tests
can be reduced by identifying the most important tests, then we could also
identify the redundant tests. By omitting the redundant tests, observation time
could be reduced and early treatment could be provided to avoid the risk.
Additionally, unnecessary monetary cost would be avoided. Our approach uses
state-ofthe- art feature selection for predicting ICU patient deterioration
using the medical lab results. We apply our technique on the publicly available
MIMIC-II database and show the effectiveness of the feature selection. We also
provide a detailed analysis of the best features identified by our approach.
|
Evaluating Prerequisite Qualities for Learning End-to-End Dialog Systems | A long-term goal of machine learning is to build intelligent conversational
agents. One recent popular approach is to train end-to-end models on a large
amount of real dialog transcripts between humans (Sordoni et al., 2015; Vinyals
& Le, 2015; Shang et al., 2015). However, this approach leaves many questions
unanswered as an understanding of the precise successes and shortcomings of
each model is hard to assess. A contrasting recent proposal are the bAbI tasks
(Weston et al., 2015b) which are synthetic data that measure the ability of
learning machines at various reasoning tasks over toy language. Unfortunately,
those tests are very small and hence may encourage methods that do not scale.
In this work, we propose a suite of new tasks of a much larger scale that
attempt to bridge the gap between the two regimes. Choosing the domain of
movies, we provide tasks that test the ability of models to answer factual
questions (utilizing OMDB), provide personalization (utilizing MovieLens),
carry short conversations about the two, and finally to perform on natural
dialogs from Reddit. We provide a dataset covering 75k movie entities and with
3.5M training examples. We present results of various models on these tasks,
and evaluate their performance.
|
Session-based Recommendations with Recurrent Neural Networks | We apply recurrent neural networks (RNN) on a new domain, namely recommender
systems. Real-life recommender systems often face the problem of having to base
recommendations only on short session-based data (e.g. a small sportsware
website) instead of long user histories (as in the case of Netflix). In this
situation the frequently praised matrix factorization approaches are not
accurate. This problem is usually overcome in practice by resorting to
item-to-item recommendations, i.e. recommending similar items. We argue that by
modeling the whole session, more accurate recommendations can be provided. We
therefore propose an RNN-based approach for session-based recommendations. Our
approach also considers practical aspects of the task and introduces several
modifications to classic RNNs such as a ranking loss function that make it more
viable for this specific problem. Experimental results on two data-sets show
marked improvements over widely used approaches.
|
Gradual DropIn of Layers to Train Very Deep Neural Networks | We introduce the concept of dynamically growing a neural network during
training. In particular, an untrainable deep network starts as a trainable
shallow network and newly added layers are slowly, organically added during
training, thereby increasing the network's depth. This is accomplished by a new
layer, which we call DropIn. The DropIn layer starts by passing the output from
a previous layer (effectively skipping over the newly added layers), then
increasingly including units from the new layers for both feedforward and
backpropagation. We show that deep networks, which are untrainable with
conventional methods, will converge with DropIn layers interspersed in the
architecture. In addition, we demonstrate that DropIn provides regularization
during training in an analogous way as dropout. Experiments are described with
the MNIST dataset and various expanded LeNet architectures, CIFAR-10 dataset
with its architecture expanded from 3 to 11 layers, and on the ImageNet dataset
with the AlexNet architecture expanded to 13 layers and the VGG 16-layer
architecture.
|
On the Linear Algebraic Structure of Distributed Word Representations | In this work, we leverage the linear algebraic structure of distributed word
representations to automatically extend knowledge bases and allow a machine to
learn new facts about the world. Our goal is to extract structured facts from
corpora in a simpler manner, without applying classifiers or patterns, and
using only the co-occurrence statistics of words. We demonstrate that the
linear algebraic structure of word embeddings can be used to reduce data
requirements for methods of learning facts. In particular, we demonstrate that
words belonging to a common category, or pairs of words satisfying a certain
relation, form a low-rank subspace in the projected space. We compute a basis
for this low-rank subspace using singular value decomposition (SVD), then use
this basis to discover new facts and to fit vectors for less frequent words
which we do not yet have vectors for.
|
Online Semi-Supervised Learning with Deep Hybrid Boltzmann Machines and
Denoising Autoencoders | Two novel deep hybrid architectures, the Deep Hybrid Boltzmann Machine and
the Deep Hybrid Denoising Auto-encoder, are proposed for handling
semi-supervised learning problems. The models combine experts that model
relevant distributions at different levels of abstraction to improve overall
predictive performance on discriminative tasks. Theoretical motivations and
algorithms for joint learning for each are presented. We apply the new models
to the domain of data-streams in work towards life-long learning. The proposed
architectures show improved performance compared to a pseudo-labeled, drop-out
rectifier network.
|
End-to-end Learning of Action Detection from Frame Glimpses in Videos | In this work we introduce a fully end-to-end approach for action detection in
videos that learns to directly predict the temporal bounds of actions. Our
intuition is that the process of detecting actions is naturally one of
observation and refinement: observing moments in video, and refining hypotheses
about when an action is occurring. Based on this insight, we formulate our
model as a recurrent neural network-based agent that interacts with a video
over time. The agent observes video frames and decides both where to look next
and when to emit a prediction. Since backpropagation is not adequate in this
non-differentiable setting, we use REINFORCE to learn the agent's decision
policy. Our model achieves state-of-the-art results on the THUMOS'14 and
ActivityNet datasets while observing only a fraction (2% or less) of the video
frames.
|
Anvaya: An Algorithm and Case-Study on Improving the Goodness of
Software Process Models generated by Mining Event-Log Data in Issue Tracking
System | Issue Tracking Systems (ITS) such as Bugzilla can be viewed as Process Aware
Information Systems (PAIS) generating event-logs during the life-cycle of a bug
report. Process Mining consists of mining event logs generated from PAIS for
process model discovery, conformance and enhancement. We apply process map
discovery techniques to mine event trace data generated from ITS of open source
Firefox browser project to generate and study process models. Bug life-cycle
consists of diversity and variance. Therefore, the process models generated
from the event-logs are spaghetti-like with large number of edges,
inter-connections and nodes. Such models are complex to analyse and difficult
to comprehend by a process analyst. We improve the Goodness (fitness and
structural complexity) of the process models by splitting the event-log into
homogeneous subsets by clustering structurally similar traces. We adapt the
K-Medoid clustering algorithm with two different distance metrics: Longest
Common Subsequence (LCS) and Dynamic Time Warping (DTW). We evaluate the
goodness of the process models generated from the clusters using complexity and
fitness metrics. We study back-forth \& self-loops, bug reopening, and
bottleneck in the clusters obtained and show that clustering enables better
analysis. We also propose an algorithm to automate the clustering process -the
algorithm takes as input the event log and returns the best cluster set.
|
Detecting Road Surface Wetness from Audio: A Deep Learning Approach | We introduce a recurrent neural network architecture for automated road
surface wetness detection from audio of tire-surface interaction. The
robustness of our approach is evaluated on 785,826 bins of audio that span an
extensive range of vehicle speeds, noises from the environment, road surface
types, and pavement conditions including international roughness index (IRI)
values from 25 in/mi to 1400 in/mi. The training and evaluation of the model
are performed on different roads to minimize the impact of environmental and
other external factors on the accuracy of the classification. We achieve an
unweighted average recall (UAR) of 93.2% across all vehicle speeds including 0
mph. The classifier still works at 0 mph because the discriminating signal is
present in the sound of other vehicles driving by.
|
ReSeg: A Recurrent Neural Network-based Model for Semantic Segmentation | We propose a structured prediction architecture, which exploits the local
generic features extracted by Convolutional Neural Networks and the capacity of
Recurrent Neural Networks (RNN) to retrieve distant dependencies. The proposed
architecture, called ReSeg, is based on the recently introduced ReNet model for
image classification. We modify and extend it to perform the more challenging
task of semantic segmentation. Each ReNet layer is composed of four RNN that
sweep the image horizontally and vertically in both directions, encoding
patches or activations, and providing relevant global information. Moreover,
ReNet layers are stacked on top of pre-trained convolutional layers, benefiting
from generic local features. Upsampling layers follow ReNet layers to recover
the original image resolution in the final predictions. The proposed ReSeg
architecture is efficient, flexible and suitable for a variety of semantic
segmentation tasks. We evaluate ReSeg on several widely-used semantic
segmentation datasets: Weizmann Horse, Oxford Flower, and CamVid; achieving
state-of-the-art performance. Results show that ReSeg can act as a suitable
architecture for semantic segmentation tasks, and may have further applications
in other structured prediction problems. The source code and model
hyperparameters are available on https://github.com/fvisin/reseg.
|
Multiple--Instance Learning: Christoffel Function Approach to
Distribution Regression Problem | A two--step Christoffel function based solution is proposed to distribution
regression problem. On the first step, to model distribution of observations
inside a bag, build Christoffel function for each bag of observations. Then, on
the second step, build outcome variable Christoffel function, but use the bag's
Christoffel function value at given point as the weight for the bag's outcome.
The approach allows the result to be obtained in closed form and then to be
evaluated numerically. While most of existing approaches minimize some kind an
error between outcome and prediction, the proposed approach is conceptually
different, because it uses Christoffel function for knowledge representation,
what is conceptually equivalent working with probabilities only. To receive
possible outcomes and their probabilities Gauss quadrature for second--step
measure can be built, then the nodes give possible outcomes and normalized
weights -- outcome probabilities. A library providing numerically stable
polynomial basis for these calculations is available, what make the proposed
approach practical.
|
On the Generalization Error Bounds of Neural Networks under
Diversity-Inducing Mutual Angular Regularization | Recently diversity-inducing regularization methods for latent variable models
(LVMs), which encourage the components in LVMs to be diverse, have been studied
to address several issues involved in latent variable modeling: (1) how to
capture long-tail patterns underlying data; (2) how to reduce model complexity
without sacrificing expressivity; (3) how to improve the interpretability of
learned patterns. While the effectiveness of diversity-inducing regularizers
such as the mutual angular regularizer has been demonstrated empirically, a
rigorous theoretical analysis of them is still missing. In this paper, we aim
to bridge this gap and analyze how the mutual angular regularizer (MAR) affects
the generalization performance of supervised LVMs. We use neural network (NN)
as a model instance to carry out the study and the analysis shows that
increasing the diversity of hidden units in NN would reduce estimation error
and increase approximation error. In addition to theoretical analysis, we also
present empirical study which demonstrates that the MAR can greatly improve the
performance of NN and the empirical observations are in accordance with the
theoretical analysis.
|
Cascading Denoising Auto-Encoder as a Deep Directed Generative Model | Recent work (Bengio et al., 2013) has shown howDenoising Auto-Encoders(DAE)
become gener-ative models as a density estimator. However,in practice, the
framework suffers from a mixingproblem in the MCMC sampling process and
nodirect method to estimate the test log-likelihood.We consider a directed
model with an stochas-tic identity mapping (simple corruption pro-cess) as an
inference model and a DAE as agenerative model. By cascading these mod-els, we
propose Cascading Denoising Auto-Encoders(CDAE) which can generate samples
ofdata distribution from tractable prior distributionunder the assumption that
probabilistic distribu-tion of corrupted data approaches tractable
priordistribution as the level of corruption increases.This work tries to
answer two questions. On theone hand, can deep directed models be success-fully
trained without intractable posterior infer-ence and difficult optimization of
very deep neu-ral networks in inference and generative mod-els? These are
unavoidable when recent suc-cessful directed model like VAE (Kingma &Welling,
2014) is trained on complex dataset likereal images. On the other hand, can
DAEs getclean samples of data distribution from heavilycorrupted samples which
can be considered oftractable prior distribution far from data mani-fold?
so-called global denoising scheme.Our results show positive responses of
thesequestions and this work can provide fairly simpleframework for generative
models of very com-plex dataset.
|
What Happened to My Dog in That Network: Unraveling Top-down Generators
in Convolutional Neural Networks | Top-down information plays a central role in human perception, but plays
relatively little role in many current state-of-the-art deep networks, such as
Convolutional Neural Networks (CNNs). This work seeks to explore a path by
which top-down information can have a direct impact within current deep
networks. We explore this path by learning and using "generators" corresponding
to the network internal effects of three types of transformation (each a
restriction of a general affine transformation): rotation, scaling, and
translation. We demonstrate how these learned generators can be used to
transfer top-down information to novel settings, as mediated by the "feature
flows" that the transformations (and the associated generators) correspond to
inside the network. Specifically, we explore three aspects: 1) using generators
as part of a method for synthesizing transformed images --- given a previously
unseen image, produce versions of that image corresponding to one or more
specified transformations, 2) "zero-shot learning" --- when provided with a
feature flow corresponding to the effect of a transformation of unknown amount,
leverage learned generators as part of a method by which to perform an accurate
categorization of the amount of transformation, even for amounts never observed
during training, and 3) (inside-CNN) "data augmentation" --- improve the
classification performance of an existing network by using the learned
generators to directly provide additional training "inside the CNN".
|
Parallel Predictive Entropy Search for Batch Global Optimization of
Expensive Objective Functions | We develop parallel predictive entropy search (PPES), a novel algorithm for
Bayesian optimization of expensive black-box objective functions. At each
iteration, PPES aims to select a batch of points which will maximize the
information gain about the global maximizer of the objective. Well known
strategies exist for suggesting a single evaluation point based on previous
observations, while far fewer are known for selecting batches of points to
evaluate in parallel. The few batch selection schemes that have been studied
all resort to greedy methods to compute an optimal batch. To the best of our
knowledge, PPES is the first non-greedy batch Bayesian optimization strategy.
We demonstrate the benefit of this approach in optimization performance on both
synthetic and real world applications, including problems in machine learning,
rocket science and robotics.
|
A PAC Approach to Application-Specific Algorithm Selection | The best algorithm for a computational problem generally depends on the
"relevant inputs," a concept that depends on the application domain and often
defies formal articulation. While there is a large literature on empirical
approaches to selecting the best algorithm for a given application domain,
there has been surprisingly little theoretical analysis of the problem.
This paper adapts concepts from statistical and online learning theory to
reason about application-specific algorithm selection. Our models capture
several state-of-the-art empirical and theoretical approaches to the problem,
ranging from self-improving algorithms to empirical performance models, and our
results identify conditions under which these approaches are guaranteed to
perform well. We present one framework that models algorithm selection as a
statistical learning problem, and our work here shows that dimension notions
from statistical learning theory, historically used to measure the complexity
of classes of binary- and real-valued functions, are relevant in a much broader
algorithmic context. We also study the online version of the algorithm
selection problem, and give possibility and impossibility results for the
existence of no-regret learning algorithms.
|
Noisy Submodular Maximization via Adaptive Sampling with Applications to
Crowdsourced Image Collection Summarization | We address the problem of maximizing an unknown submodular function that can
only be accessed via noisy evaluations. Our work is motivated by the task of
summarizing content, e.g., image collections, by leveraging users' feedback in
form of clicks or ratings. For summarization tasks with the goal of maximizing
coverage and diversity, submodular set functions are a natural choice. When the
underlying submodular function is unknown, users' feedback can provide noisy
evaluations of the function that we seek to maximize. We provide a generic
algorithm -- \submM{} -- for maximizing an unknown submodular function under
cardinality constraints. This algorithm makes use of a novel exploration module
-- \blbox{} -- that proposes good elements based on adaptively sampling noisy
function evaluations. \blbox{} is able to accommodate different kinds of
observation models such as value queries and pairwise comparisons. We provide
PAC-style guarantees on the quality and sampling cost of the solution obtained
by \submM{}. We demonstrate the effectiveness of our approach in an
interactive, crowdsourced image collection summarization application.
|
NetVLAD: CNN architecture for weakly supervised place recognition | We tackle the problem of large scale visual place recognition, where the task
is to quickly and accurately recognize the location of a given query
photograph. We present the following three principal contributions. First, we
develop a convolutional neural network (CNN) architecture that is trainable in
an end-to-end manner directly for the place recognition task. The main
component of this architecture, NetVLAD, is a new generalized VLAD layer,
inspired by the "Vector of Locally Aggregated Descriptors" image representation
commonly used in image retrieval. The layer is readily pluggable into any CNN
architecture and amenable to training via backpropagation. Second, we develop a
training procedure, based on a new weakly supervised ranking loss, to learn
parameters of the architecture in an end-to-end manner from images depicting
the same places over time downloaded from Google Street View Time Machine.
Finally, we show that the proposed architecture significantly outperforms
non-learnt image representations and off-the-shelf CNN descriptors on two
challenging place recognition benchmarks, and improves over current
state-of-the-art compact image representations on standard image retrieval
benchmarks.
|
Input Sparsity Time Low-Rank Approximation via Ridge Leverage Score
Sampling | We present a new algorithm for finding a near optimal low-rank approximation
of a matrix $A$ in $O(nnz(A))$ time. Our method is based on a recursive
sampling scheme for computing a representative subset of $A$'s columns, which
is then used to find a low-rank approximation.
This approach differs substantially from prior $O(nnz(A))$ time algorithms,
which are all based on fast Johnson-Lindenstrauss random projections. It
matches the guarantees of these methods while offering a number of advantages.
Not only are sampling algorithms faster for sparse and structured data, but
they can also be applied in settings where random projections cannot. For
example, we give new single-pass streaming algorithms for the column subset
selection and projection-cost preserving sample problems. Our method has also
been used to give the fastest algorithms for provably approximating kernel
matrices [MM16].
|
Learning Simple Algorithms from Examples | We present an approach for learning simple algorithms such as copying,
multi-digit addition and single digit multiplication directly from examples.
Our framework consists of a set of interfaces, accessed by a controller.
Typical interfaces are 1-D tapes or 2-D grids that hold the input and output
data. For the controller, we explore a range of neural network-based models
which vary in their ability to abstract the underlying algorithm from training
instances and generalize to test examples with many thousands of digits. The
controller is trained using $Q$-learning with several enhancements and we show
that the bottleneck is in the capabilities of the controller rather than in the
search incurred by $Q$-learning.
|
Fast and Accurate Deep Network Learning by Exponential Linear Units
(ELUs) | We introduce the "exponential linear unit" (ELU) which speeds up learning in
deep neural networks and leads to higher classification accuracies. Like
rectified linear units (ReLUs), leaky ReLUs (LReLUs) and parametrized ReLUs
(PReLUs), ELUs alleviate the vanishing gradient problem via the identity for
positive values. However, ELUs have improved learning characteristics compared
to the units with other activation functions. In contrast to ReLUs, ELUs have
negative values which allows them to push mean unit activations closer to zero
like batch normalization but with lower computational complexity. Mean shifts
toward zero speed up learning by bringing the normal gradient closer to the
unit natural gradient because of a reduced bias shift effect. While LReLUs and
PReLUs have negative values, too, they do not ensure a noise-robust
deactivation state. ELUs saturate to a negative value with smaller inputs and
thereby decrease the forward propagated variation and information. Therefore,
ELUs code the degree of presence of particular phenomena in the input, while
they do not quantitatively model the degree of their absence. In experiments,
ELUs lead not only to faster learning, but also to significantly better
generalization performance than ReLUs and LReLUs on networks with more than 5
layers. On CIFAR-100 ELUs networks significantly outperform ReLU networks with
batch normalization while batch normalization does not improve ELU networks.
ELU networks are among the top 10 reported CIFAR-10 results and yield the best
published result on CIFAR-100, without resorting to multi-view evaluation or
model averaging. On ImageNet, ELU networks considerably speed up learning
compared to a ReLU network with the same architecture, obtaining less than 10%
classification error for a single crop, single model network.
|
Sparse Recovery via Partial Regularization: Models, Theory and
Algorithms | In the context of sparse recovery, it is known that most of existing
regularizers such as $\ell_1$ suffer from some bias incurred by some leading
entries (in magnitude) of the associated vector. To neutralize this bias, we
propose a class of models with partial regularizers for recovering a sparse
solution of a linear system. We show that every local minimizer of these models
is sufficiently sparse or the magnitude of all its nonzero entries is above a
uniform constant depending only on the data of the linear system. Moreover, for
a class of partial regularizers, any global minimizer of these models is a
sparsest solution to the linear system. We also establish some sufficient
conditions for local or global recovery of the sparsest solution to the linear
system, among which one of the conditions is weaker than the best known
restricted isometry property (RIP) condition for sparse recovery by $\ell_1$.
In addition, a first-order feasible augmented Lagrangian (FAL) method is
proposed for solving these models, in which each subproblem is solved by a
nonmonotone proximal gradient (NPG) method. Despite the complication of the
partial regularizers, we show that each proximal subproblem in NPG can be
solved as a certain number of one-dimensional optimization problems, which
usually have a closed-form solution. We also show that any accumulation point
of the sequence generated by FAL is a first-order stationary point of the
models. Numerical results on compressed sensing and sparse logistic regression
demonstrate that the proposed models substantially outperform the widely used
ones in the literature in terms of solution quality.
|
Modular Autoencoders for Ensemble Feature Extraction | We introduce the concept of a Modular Autoencoder (MAE), capable of learning
a set of diverse but complementary representations from unlabelled data, that
can later be used for supervised tasks. The learning of the representations is
controlled by a trade off parameter, and we show on six benchmark datasets the
optimum lies between two extremes: a set of smaller, independent autoencoders
each with low capacity, versus a single monolithic encoding, outperforming an
appropriate baseline. In the present paper we explore the special case of
linear MAE, and derive an SVD-based algorithm which converges several orders of
magnitude faster than gradient descent.
|
Interpretable Two-level Boolean Rule Learning for Classification | This paper proposes algorithms for learning two-level Boolean rules in
Conjunctive Normal Form (CNF, i.e. AND-of-ORs) or Disjunctive Normal Form (DNF,
i.e. OR-of-ANDs) as a type of human-interpretable classification model, aiming
for a favorable trade-off between the classification accuracy and the
simplicity of the rule. Two formulations are proposed. The first is an integer
program whose objective function is a combination of the total number of errors
and the total number of features used in the rule. We generalize a previously
proposed linear programming (LP) relaxation from one-level to two-level rules.
The second formulation replaces the 0-1 classification error with the Hamming
distance from the current two-level rule to the closest rule that correctly
classifies a sample. Based on this second formulation, block coordinate descent
and alternating minimization algorithms are developed. Experiments show that
the two-level rules can yield noticeably better performance than one-level
rules due to their dramatically larger modeling capacity, and the two
algorithms based on the Hamming distance formulation are generally superior to
the other two-level rule learning methods in our comparison. A proposed
approach to binarize any fractional values in the optimal solutions of LP
relaxations is also shown to be effective.
|
Pushing the Boundaries of Boundary Detection using Deep Learning | In this work we show that adapting Deep Convolutional Neural Network training
to the task of boundary detection can result in substantial improvements over
the current state-of-the-art in boundary detection.
Our contributions consist firstly in combining a careful design of the loss
for boundary detection training, a multi-resolution architecture and training
with external data to improve the detection accuracy of the current state of
the art. When measured on the standard Berkeley Segmentation Dataset, we
improve theoptimal dataset scale F-measure from 0.780 to 0.808 - while human
performance is at 0.803. We further improve performance to 0.813 by combining
deep learning with grouping, integrating the Normalized Cuts technique within a
deep network.
We also examine the potential of our boundary detector in conjunction with
the task of semantic segmentation and demonstrate clear improvements over
state-of-the-art systems. Our detector is fully integrated in the popular Caffe
framework and processes a 320x420 image in less than a second.
|
MazeBase: A Sandbox for Learning from Games | This paper introduces MazeBase: an environment for simple 2D games, designed
as a sandbox for machine learning approaches to reasoning and planning. Within
it, we create 10 simple games embodying a range of algorithmic tasks (e.g.
if-then statements or set negation). A variety of neural models (fully
connected, convolutional network, memory network) are deployed via
reinforcement learning on these games, with and without a procedurally
generated curriculum. Despite the tasks' simplicity, the performance of the
models is far from optimal, suggesting directions for future development. We
also demonstrate the versatility of MazeBase by using it to emulate small
combat scenarios from StarCraft. Models trained on the MazeBase version can be
directly applied to StarCraft, where they consistently beat the in-game AI.
|
Top-Down Learning for Structured Labeling with Convolutional Pseudoprior | Current practice in convolutional neural networks (CNN) remains largely
bottom-up and the role of top-down process in CNN for pattern analysis and
visual inference is not very clear. In this paper, we propose a new method for
structured labeling by developing convolutional pseudo-prior (ConvPP) on the
ground-truth labels. Our method has several interesting properties: (1)
compared with classical machine learning algorithms like CRFs and Structural
SVM, ConvPP automatically learns rich convolutional kernels to capture both
short- and long- range contexts; (2) compared with cascade classifiers like
Auto-Context, ConvPP avoids the iterative steps of learning a series of
discriminative classifiers and automatically learns contextual configurations;
(3) compared with recent efforts combing CNN models with CRFs and RNNs, ConvPP
learns convolution in the labeling space with much improved modeling capability
and less manual specification; (4) compared with Bayesian models like MRFs,
ConvPP capitalizes on the rich representation power of convolution by
automatically learning priors built on convolutional filters. We accomplish our
task using pseudo-likelihood approximation to the prior under a novel
fixed-point network structure that facilitates an end-to-end learning process.
We show state-of-the-art results on sequential labeling and image labeling
benchmarks.
|
Weak Convergence Properties of Constrained Emphatic Temporal-difference
Learning with Constant and Slowly Diminishing Stepsize | We consider the emphatic temporal-difference (TD) algorithm, ETD($\lambda$),
for learning the value functions of stationary policies in a discounted, finite
state and action Markov decision process. The ETD($\lambda$) algorithm was
recently proposed by Sutton, Mahmood, and White to solve a long-standing
divergence problem of the standard TD algorithm when it is applied to
off-policy training, where data from an exploratory policy are used to evaluate
other policies of interest. The almost sure convergence of ETD($\lambda$) has
been proved in our recent work under general off-policy training conditions,
but for a narrow range of diminishing stepsize. In this paper we present
convergence results for constrained versions of ETD($\lambda$) with constant
stepsize and with diminishing stepsize from a broad range. Our results
characterize the asymptotic behavior of the trajectory of iterates produced by
those algorithms, and are derived by combining key properties of ETD($\lambda$)
with powerful convergence theorems from the weak convergence methods in
stochastic approximation theory. For the case of constant stepsize, in addition
to analyzing the behavior of the algorithms in the limit as the stepsize
parameter approaches zero, we also analyze their behavior for a fixed stepsize
and bound the deviations of their averaged iterates from the desired solution.
These results are obtained by exploiting the weak Feller property of the Markov
chains associated with the algorithms, and by using ergodic theorems for weak
Feller Markov chains, in conjunction with the convergence results we get from
the weak convergence methods. Besides ETD($\lambda$), our analysis also applies
to the off-policy TD($\lambda$) algorithm, when the divergence issue is avoided
by setting $\lambda$ sufficiently large.
|
Constrained Structured Regression with Convolutional Neural Networks | Convolutional Neural Networks (CNNs) have recently emerged as the dominant
model in computer vision. If provided with enough training data, they predict
almost any visual quantity. In a discrete setting, such as classification, CNNs
are not only able to predict a label but often predict a confidence in the form
of a probability distribution over the output space. In continuous regression
tasks, such a probability estimate is often lacking. We present a regression
framework which models the output distribution of neural networks. This output
distribution allows us to infer the most likely labeling following a set of
physical or modeling constraints. These constraints capture the intricate
interplay between different input and output variables, and complement the
output of a CNN. However, they may not hold everywhere. Our setup further
allows to learn a confidence with which a constraint holds, in the form of a
distribution of the constrain satisfaction. We evaluate our approach on the
problem of intrinsic image decomposition, and show that constrained structured
regression significantly increases the state-of-the-art.
|
The Limitations of Deep Learning in Adversarial Settings | Deep learning takes advantage of large datasets and computationally efficient
training algorithms to outperform other approaches at various machine learning
tasks. However, imperfections in the training phase of deep neural networks
make them vulnerable to adversarial samples: inputs crafted by adversaries with
the intent of causing deep neural networks to misclassify. In this work, we
formalize the space of adversaries against deep neural networks (DNNs) and
introduce a novel class of algorithms to craft adversarial samples based on a
precise understanding of the mapping between inputs and outputs of DNNs. In an
application to computer vision, we show that our algorithms can reliably
produce samples correctly classified by human subjects but misclassified in
specific targets by a DNN with a 97% adversarial success rate while only
modifying on average 4.02% of the input features per sample. We then evaluate
the vulnerability of different sample classes to adversarial perturbations by
defining a hardness measure. Finally, we describe preliminary work outlining
defenses against adversarial samples by defining a predictive measure of
distance between a benign input and a target classification.
|
Convergent Learning: Do different neural networks learn the same
representations? | Recent success in training deep neural networks have prompted active
investigation into the features learned on their intermediate layers. Such
research is difficult because it requires making sense of non-linear
computations performed by millions of parameters, but valuable because it
increases our ability to understand current models and create improved versions
of them. In this paper we investigate the extent to which neural networks
exhibit what we call convergent learning, which is when the representations
learned by multiple nets converge to a set of features which are either
individually similar between networks or where subsets of features span similar
low-dimensional spaces. We propose a specific method of probing
representations: training multiple networks and then comparing and contrasting
their individual, learned representations at the level of neurons or groups of
neurons. We begin research into this question using three techniques to
approximately align different neural networks on a feature level: a bipartite
matching approach that makes one-to-one assignments between neurons, a sparse
prediction approach that finds one-to-many mappings, and a spectral clustering
approach that finds many-to-many mappings. This initial investigation reveals a
few previously unknown properties of neural networks, and we argue that future
research into the question of convergent learning will yield many more. The
insights described here include (1) that some features are learned reliably in
multiple networks, yet other features are not consistently learned; (2) that
units learn to span low-dimensional subspaces and, while these subspaces are
common to multiple networks, the specific basis vectors learned are not; (3)
that the representation codes show evidence of being a mix between a local code
and slightly, but not fully, distributed codes across multiple units.
|
Transductive Log Opinion Pool of Gaussian Process Experts | We introduce a framework for analyzing transductive combination of Gaussian
process (GP) experts, where independently trained GP experts are combined in a
way that depends on test point location, in order to scale GPs to big data. The
framework provides some theoretical justification for the generalized product
of GP experts (gPoE-GP) which was previously shown to work well in practice but
lacks theoretical basis. Based on the proposed framework, an improvement over
gPoE-GP is introduced and empirically validated.
|
DenseCap: Fully Convolutional Localization Networks for Dense Captioning | We introduce the dense captioning task, which requires a computer vision
system to both localize and describe salient regions in images in natural
language. The dense captioning task generalizes object detection when the
descriptions consist of a single word, and Image Captioning when one predicted
region covers the full image. To address the localization and description task
jointly we propose a Fully Convolutional Localization Network (FCLN)
architecture that processes an image with a single, efficient forward pass,
requires no external regions proposals, and can be trained end-to-end with a
single round of optimization. The architecture is composed of a Convolutional
Network, a novel dense localization layer, and Recurrent Neural Network
language model that generates the label sequences. We evaluate our network on
the Visual Genome dataset, which comprises 94,000 images and 4,100,000
region-grounded captions. We observe both speed and accuracy improvements over
baselines based on current state of the art approaches in both generation and
retrieval settings.
|
Picking a Conveyor Clean by an Autonomously Learning Robot | We present a research picking prototype related to our company's industrial
waste sorting application. The goal of the prototype is to be as autonomous as
possible and it both calibrates itself and improves its picking with minimal
human intervention. The system learns to pick objects better based on a
feedback sensor in its gripper and uses machine learning to choosing the best
proposal from a random sample produced by simple hard-coded geometric models.
We show experimentally the system improving its picking autonomously by
measuring the pick success rate as function of time. We also show how this
system can pick a conveyor belt clean, depositing 70 out of 80 objects in a
difficult to manipulate pile of novel objects into the correct chute. We
discuss potential improvements and next steps in this direction.
|
LocNet: Improving Localization Accuracy for Object Detection | We propose a novel object localization methodology with the purpose of
boosting the localization accuracy of state-of-the-art object detection
systems. Our model, given a search region, aims at returning the bounding box
of an object of interest inside this region. To accomplish its goal, it relies
on assigning conditional probabilities to each row and column of this region,
where these probabilities provide useful information regarding the location of
the boundaries of the object inside the search region and allow the accurate
inference of the object bounding box under a simple probabilistic framework.
For implementing our localization model, we make use of a convolutional
neural network architecture that is properly adapted for this task, called
LocNet. We show experimentally that LocNet achieves a very significant
improvement on the mAP for high IoU thresholds on PASCAL VOC2007 test set and
that it can be very easily coupled with recent state-of-the-art object
detection systems, helping them to boost their performance. Finally, we
demonstrate that our detection approach can achieve high detection accuracy
even when it is given as input a set of sliding windows, thus proving that it
is independent of box proposal methods.
|
Generalized Conjugate Gradient Methods for $\ell_1$ Regularized Convex
Quadratic Programming with Finite Convergence | The conjugate gradient (CG) method is an efficient iterative method for
solving large-scale strongly convex quadratic programming (QP). In this paper
we propose some generalized CG (GCG) methods for solving the
$\ell_1$-regularized (possibly not strongly) convex QP that terminate at an
optimal solution in a finite number of iterations. At each iteration, our
methods first identify a face of an orthant and then either perform an exact
line search along the direction of the negative projected minimum-norm
subgradient of the objective function or execute a CG subroutine that conducts
a sequence of CG iterations until a CG iterate crosses the boundary of this
face or an approximate minimizer of over this face or a subface is found. We
determine which type of step should be taken by comparing the magnitude of some
components of the minimum-norm subgradient of the objective function to that of
its rest components. Our analysis on finite convergence of these methods makes
use of an error bound result and some key properties of the aforementioned
exact line search and the CG subroutine. We also show that the proposed methods
are capable of finding an approximate solution of the problem by allowing some
inexactness on the execution of the CG subroutine. The overall arithmetic
operation cost of our GCG methods for finding an $\epsilon$-optimal solution
depends on $\epsilon$ in $O(\log(1/\epsilon))$, which is superior to the
accelerated proximal gradient method [2,23] that depends on $\epsilon$ in
$O(1/\sqrt{\epsilon})$. In addition, our GCG methods can be extended
straightforwardly to solve box-constrained convex QP with finite convergence.
Numerical results demonstrate that our methods are very favorable for solving
ill-conditioned problems.
|
Dynamic Capacity Networks | We introduce the Dynamic Capacity Network (DCN), a neural network that can
adaptively assign its capacity across different portions of the input data.
This is achieved by combining modules of two types: low-capacity sub-networks
and high-capacity sub-networks. The low-capacity sub-networks are applied
across most of the input, but also provide a guide to select a few portions of
the input on which to apply the high-capacity sub-networks. The selection is
made using a novel gradient-based attention mechanism, that efficiently
identifies input regions for which the DCN's output is most sensitive and to
which we should devote more capacity. We focus our empirical evaluation on the
Cluttered MNIST and SVHN image datasets. Our findings indicate that DCNs are
able to drastically reduce the number of computations, compared to traditional
convolutional neural networks, while maintaining similar or even better
performance.
|
Private Posterior distributions from Variational approximations | Privacy preserving mechanisms such as differential privacy inject additional
randomness in the form of noise in the data, beyond the sampling mechanism.
Ignoring this additional noise can lead to inaccurate and invalid inferences.
In this paper, we incorporate the privacy mechanism explicitly into the
likelihood function by treating the original data as missing, with an end goal
of estimating posterior distributions over model parameters. This leads to a
principled way of performing valid statistical inference using private data,
however, the corresponding likelihoods are intractable. In this paper, we
derive fast and accurate variational approximations to tackle such intractable
likelihoods that arise due to privacy. We focus on estimating posterior
distributions of parameters of the naive Bayes log-linear model, where the
sufficient statistics of this model are shared using a differentially private
interface. Using a simulation study, we show that the posterior approximations
outperform the naive method of ignoring the noise addition mechanism.
|
Performance Limits of Stochastic Sub-Gradient Learning, Part I: Single
Agent Case | In this work and the supporting Part II, we examine the performance of
stochastic sub-gradient learning strategies under weaker conditions than
usually considered in the literature. The new conditions are shown to be
automatically satisfied by several important cases of interest including SVM,
LASSO, and Total-Variation denoising formulations. In comparison, these
problems do not satisfy the traditional assumptions used in prior analyses and,
therefore, conclusions derived from these earlier treatments are not directly
applicable to these problems. The results in this article establish that
stochastic sub-gradient strategies can attain linear convergence rates, as
opposed to sub-linear rates, to the steady-state regime. A realizable
exponential-weighting procedure is employed to smooth the intermediate iterates
and guarantee useful performance bounds in terms of convergence rate and
excessive risk performance. Part I of this work focuses on single-agent
scenarios, which are common in stand-alone learning applications, while Part II
extends the analysis to networked learners. The theoretical conclusions are
illustrated by several examples and simulations, including comparisons with the
FISTA procedure.
|
Context-aware CNNs for person head detection | Person detection is a key problem for many computer vision tasks. While face
detection has reached maturity, detecting people under a full variation of
camera view-points, human poses, lighting conditions and occlusions is still a
difficult challenge. In this work we focus on detecting human heads in natural
scenes. Starting from the recent local R-CNN object detector, we extend it with
two types of contextual cues. First, we leverage person-scene relations and
propose a Global CNN model trained to predict positions and scales of heads
directly from the full image. Second, we explicitly model pairwise relations
among objects and train a Pairwise CNN model using a structured-output
surrogate loss. The Local, Global and Pairwise models are combined into a joint
CNN framework. To train and test our full model, we introduce a large dataset
composed of 369,846 human heads annotated in 224,740 movie frames. We evaluate
our method and demonstrate improvements of person head detection against
several recent baselines in three datasets. We also show improvements of the
detection speed provided by our model.
|
Temporal Convolutional Neural Networks for Diagnosis from Lab Tests | Early diagnosis of treatable diseases is essential for improving healthcare,
and many diseases' onsets are predictable from annual lab tests and their
temporal trends. We introduce a multi-resolution convolutional neural network
for early detection of multiple diseases from irregularly measured sparse lab
values. Our novel architecture takes as input both an imputed version of the
data and a binary observation matrix. For imputing the temporal sparse
observations, we develop a flexible, fast to train method for differentiable
multivariate kernel regression. Our experiments on data from 298K individuals
over 8 years, 18 common lab measurements, and 171 diseases show that the
temporal signatures learned via convolution are significantly more predictive
than baselines commonly used for early disease diagnosis.
|
Learning Halfspaces and Neural Networks with Random Initialization | We study non-convex empirical risk minimization for learning halfspaces and
neural networks. For loss functions that are $L$-Lipschitz continuous, we
present algorithms to learn halfspaces and multi-layer neural networks that
achieve arbitrarily small excess risk $\epsilon>0$. The time complexity is
polynomial in the input dimension $d$ and the sample size $n$, but exponential
in the quantity $(L/\epsilon^2)\log(L/\epsilon)$. These algorithms run multiple
rounds of random initialization followed by arbitrary optimization steps. We
further show that if the data is separable by some neural network with constant
margin $\gamma>0$, then there is a polynomial-time algorithm for learning a
neural network that separates the training data with margin $\Omega(\gamma)$.
As a consequence, the algorithm achieves arbitrary generalization error
$\epsilon>0$ with ${\rm poly}(d,1/\epsilon)$ sample and time complexity. We
establish the same learnability result when the labels are randomly flipped
with probability $\eta<1/2$.
|
Exploring Correlation between Labels to improve Multi-Label
Classification | This paper attempts multi-label classification by extending the idea of
independent binary classification models for each output label, and exploring
how the inherent correlation between output labels can be used to improve
predictions. Logistic Regression, Naive Bayes, Random Forest, and SVM models
were constructed, with SVM giving the best results: an improvement of 12.9\%
over binary models was achieved for hold out cross validation by augmenting
with pairwise correlation probabilities of the labels.
|
MOOCs Meet Measurement Theory: A Topic-Modelling Approach | This paper adapts topic models to the psychometric testing of MOOC students
based on their online forum postings. Measurement theory from education and
psychology provides statistical models for quantifying a person's attainment of
intangible attributes such as attitudes, abilities or intelligence. Such models
infer latent skill levels by relating them to individuals' observed responses
on a series of items such as quiz questions. The set of items can be used to
measure a latent skill if individuals' responses on them conform to a Guttman
scale. Such well-scaled items differentiate between individuals and inferred
levels span the entire range from most basic to the advanced. In practice,
education researchers manually devise items (quiz questions) while optimising
well-scaled conformance. Due to the costly nature and expert requirements of
this process, psychometric testing has found limited use in everyday teaching.
We aim to develop usable measurement models for highly-instrumented MOOC
delivery platforms, by using participation in automatically-extracted online
forum topics as items. The challenge is to formalise the Guttman scale
educational constraint and incorporate it into topic models. To favour topics
that automatically conform to a Guttman scale, we introduce a novel
regularisation into non-negative matrix factorisation-based topic modelling. We
demonstrate the suitability of our approach with both quantitative experiments
on three Coursera MOOCs, and with a qualitative survey of topic
interpretability on two MOOCs by domain expert interviews.
|
Learning with Memory Embeddings | Embedding learning, a.k.a. representation learning, has been shown to be able
to model large-scale semantic knowledge graphs. A key concept is a mapping of
the knowledge graph to a tensor representation whose entries are predicted by
models using latent representations of generalized entities. Latent variable
models are well suited to deal with the high dimensionality and sparsity of
typical knowledge graphs. In recent publications the embedding models were
extended to also consider time evolutions, time patterns and subsymbolic
representations. In this paper we map embedding models, which were developed
purely as solutions to technical problems for modelling temporal knowledge
graphs, to various cognitive memory functions, in particular to semantic and
concept memory, episodic memory, sensory memory, short-term memory, and working
memory. We discuss learning, query answering, the path from sensory input to
semantic decoding, and the relationship between episodic memory and semantic
memory. We introduce a number of hypotheses on human memory that can be derived
from the developed mathematical models.
|
Learning to detect video events from zero or very few video examples | In this work we deal with the problem of high-level event detection in video.
Specifically, we study the challenging problems of i) learning to detect video
events from solely a textual description of the event, without using any
positive video examples, and ii) additionally exploiting very few positive
training samples together with a small number of ``related'' videos. For
learning only from an event's textual description, we first identify a general
learning framework and then study the impact of different design choices for
various stages of this framework. For additionally learning from example
videos, when true positive training samples are scarce, we employ an extension
of the Support Vector Machine that allows us to exploit ``related'' event
videos by automatically introducing different weights for subsets of the videos
in the overall training set. Experimental evaluations performed on the
large-scale TRECVID MED 2014 video dataset provide insight on the effectiveness
of the proposed methods.
|
Relaxed Majorization-Minimization for Non-smooth and Non-convex
Optimization | We propose a new majorization-minimization (MM) method for non-smooth and
non-convex programs, which is general enough to include the existing MM
methods. Besides the local majorization condition, we only require that the
difference between the directional derivatives of the objective function and
its surrogate function vanishes when the number of iterations approaches
infinity, which is a very weak condition. So our method can use a surrogate
function that directly approximates the non-smooth objective function. In
comparison, all the existing MM methods construct the surrogate function by
approximating the smooth component of the objective function. We apply our
relaxed MM methods to the robust matrix factorization (RMF) problem with
different regularizations, where our locally majorant algorithm shows
advantages over the state-of-the-art approaches for RMF. This is the first
algorithm for RMF ensuring, without extra assumptions, that any limit point of
the iterates is a stationary point.
|
Strategic Dialogue Management via Deep Reinforcement Learning | Artificially intelligent agents equipped with strategic skills that can
negotiate during their interactions with other natural or artificial agents are
still underdeveloped. This paper describes a successful application of Deep
Reinforcement Learning (DRL) for training intelligent agents with strategic
conversational skills, in a situated dialogue setting. Previous studies have
modelled the behaviour of strategic agents using supervised learning and
traditional reinforcement learning techniques, the latter using tabular
representations or learning with linear function approximation. In this study,
we apply DRL with a high-dimensional state space to the strategic board game of
Settlers of Catan---where players can offer resources in exchange for others
and they can also reply to offers made by other players. Our experimental
results report that the DRL-based learnt policies significantly outperformed
several baselines including random, rule-based, and supervised-based
behaviours. The DRL-based policy has a 53% win rate versus 3 automated players
(`bots'), whereas a supervised player trained on a dialogue corpus in this
setting achieved only 27%, versus the same 3 bots. This result supports the
claim that DRL is a promising framework for training dialogue systems, and
strategic agents with negotiation abilities.
|
Unifying Decision Trees Split Criteria Using Tsallis Entropy | The construction of efficient and effective decision trees remains a key
topic in machine learning because of their simplicity and flexibility. A lot of
heuristic algorithms have been proposed to construct near-optimal decision
trees. ID3, C4.5 and CART are classical decision tree algorithms and the split
criteria they used are Shannon entropy, Gain Ratio and Gini index respectively.
All the split criteria seem to be independent, actually, they can be unified in
a Tsallis entropy framework. Tsallis entropy is a generalization of Shannon
entropy and provides a new approach to enhance decision trees' performance with
an adjustable parameter $q$. In this paper, a Tsallis Entropy Criterion (TEC)
algorithm is proposed to unify Shannon entropy, Gain Ratio and Gini index,
which generalizes the split criteria of decision trees. More importantly, we
reveal the relations between Tsallis entropy with different $q$ and other split
criteria. Experimental results on UCI data sets indicate that the TEC algorithm
achieves statistically significant improvement over the classical algorithms.
|
Towards Universal Paraphrastic Sentence Embeddings | We consider the problem of learning general-purpose, paraphrastic sentence
embeddings based on supervision from the Paraphrase Database (Ganitkevitch et
al., 2013). We compare six compositional architectures, evaluating them on
annotated textual similarity datasets drawn both from the same distribution as
the training data and from a wide range of other domains. We find that the most
complex architectures, such as long short-term memory (LSTM) recurrent neural
networks, perform best on the in-domain data. However, in out-of-domain
scenarios, simple architectures such as word averaging vastly outperform LSTMs.
Our simplest averaging model is even competitive with systems tuned for the
particular tasks while also being extremely efficient and easy to use.
In order to better understand how these architectures compare, we conduct
further experiments on three supervised NLP tasks: sentence similarity,
entailment, and sentiment classification. We again find that the word averaging
models perform well for sentence similarity and entailment, outperforming
LSTMs. However, on sentiment classification, we find that the LSTM performs
very strongly-even recording new state-of-the-art performance on the Stanford
Sentiment Treebank.
We then demonstrate how to combine our pretrained sentence embeddings with
these supervised tasks, using them both as a prior and as a black box feature
extractor. This leads to performance rivaling the state of the art on the SICK
similarity and entailment tasks. We release all of our resources to the
research community with the hope that they can serve as the new baseline for
further work on universal sentence embeddings.
|
Neural GPUs Learn Algorithms | Learning an algorithm from examples is a fundamental problem that has been
widely studied. Recently it has been addressed using neural networks, in
particular by Neural Turing Machines (NTMs). These are fully differentiable
computers that use backpropagation to learn their own programming. Despite
their appeal NTMs have a weakness that is caused by their sequential nature:
they are not parallel and are are hard to train due to their large depth when
unfolded.
We present a neural network architecture to address this problem: the Neural
GPU. It is based on a type of convolutional gated recurrent unit and, like the
NTM, is computationally universal. Unlike the NTM, the Neural GPU is highly
parallel which makes it easier to train and efficient to run.
An essential property of algorithms is their ability to handle inputs of
arbitrary size. We show that the Neural GPU can be trained on short instances
of an algorithmic task and successfully generalize to long instances. We
verified it on a number of tasks including long addition and long
multiplication of numbers represented in binary. We train the Neural GPU on
numbers with upto 20 bits and observe no errors whatsoever while testing it,
even on much longer numbers.
To achieve these results we introduce a technique for training deep recurrent
networks: parameter sharing relaxation. We also found a small amount of dropout
and gradient noise to have a large positive effect on learning and
generalization.
|
Hierarchical classification of e-commerce related social media | In this paper, we attempt to classify tweets into root categories of the
Amazon browse node hierarchy using a set of tweets with browse node ID labels,
a much larger set of tweets without labels, and a set of Amazon reviews.
Examining twitter data presents unique challenges in that the samples are short
(under 140 characters) and often contain misspellings or abbreviations that are
trivial for a human to decipher but difficult for a computer to parse. A
variety of query and document expansion techniques are implemented in an effort
to improve information retrieval to modest success.
|
Named Entity Recognition with Bidirectional LSTM-CNNs | Named entity recognition is a challenging task that has traditionally
required large amounts of knowledge in the form of feature engineering and
lexicons to achieve high performance. In this paper, we present a novel neural
network architecture that automatically detects word- and character-level
features using a hybrid bidirectional LSTM and CNN architecture, eliminating
the need for most feature engineering. We also propose a novel method of
encoding partial lexicon matches in neural networks and compare it to existing
approaches. Extensive evaluation shows that, given only tokenized text and
publicly available word embeddings, our system is competitive on the CoNLL-2003
dataset and surpasses the previously reported state of the art performance on
the OntoNotes 5.0 dataset by 2.13 F1 points. By using two lexicons constructed
from publicly-available sources, we establish new state of the art performance
with an F1 score of 91.62 on CoNLL-2003 and 86.28 on OntoNotes, surpassing
systems that employ heavy feature engineering, proprietary lexicons, and rich
entity linking information.
|
Random Forests for Big Data | Big Data is one of the major challenges of statistical science and has
numerous consequences from algorithmic and theoretical viewpoints. Big Data
always involve massive data but they also often include online data and data
heterogeneity. Recently some statistical methods have been adapted to process
Big Data, like linear regression models, clustering methods and bootstrapping
schemes. Based on decision trees combined with aggregation and bootstrap ideas,
random forests were introduced by Breiman in 2001. They are a powerful
nonparametric statistical method allowing to consider in a single and versatile
framework regression problems, as well as two-class and multi-class
classification problems. Focusing on classification problems, this paper
proposes a selective review of available proposals that deal with scaling
random forests to Big Data problems. These proposals rely on parallel
environments or on online adaptations of random forests. We also describe how
related quantities -- such as out-of-bag error and variable importance -- are
addressed in these methods. Then, we formulate various remarks for random
forests in the Big Data context. Finally, we experiment five variants on two
massive datasets (15 and 120 millions of observations), a simulated one as well
as real world data. One variant relies on subsampling while three others are
related to parallel implementations of random forests and involve either
various adaptations of bootstrap to Big Data or to "divide-and-conquer"
approaches. The fifth variant relates on online learning of random forests.
These numerical experiments lead to highlight the relative performance of the
different variants, as well as some of their limitations.
|
The Automatic Statistician: A Relational Perspective | Gaussian Processes (GPs) provide a general and analytically tractable way of
modeling complex time-varying, nonparametric functions. The Automatic Bayesian
Covariance Discovery (ABCD) system constructs natural-language description of
time-series data by treating unknown time-series data nonparametrically using
GP with a composite covariance kernel function. Unfortunately, learning a
composite covariance kernel with a single time-series data set often results in
less informative kernel that may not give qualitative, distinctive descriptions
of data. We address this challenge by proposing two relational kernel learning
methods which can model multiple time-series data sets by finding common,
shared causes of changes. We show that the relational kernel learning methods
find more accurate models for regression problems on several real-world data
sets; US stock data, US house price index data and currency exchange rate data.
|
Regularizing RNNs by Stabilizing Activations | We stabilize the activations of Recurrent Neural Networks (RNNs) by
penalizing the squared distance between successive hidden states' norms.
This penalty term is an effective regularizer for RNNs including LSTMs and
IRNNs, improving performance on character-level language modeling and phoneme
recognition, and outperforming weight noise and dropout.
We achieve competitive performance (18.6\% PER) on the TIMIT phoneme
recognition task for RNNs evaluated without beam search or an RNN transducer.
With this penalty term, IRNN can achieve similar performance to LSTM on
language modeling, although adding the penalty term to the LSTM results in
superior performance.
Our penalty term also prevents the exponential growth of IRNN's activations
outside of their training horizon, allowing them to generalize to much longer
sequences.
|
Gains and Losses are Fundamentally Different in Regret Minimization: The
Sparse Case | We demonstrate that, in the classical non-stochastic regret minimization
problem with $d$ decisions, gains and losses to be respectively maximized or
minimized are fundamentally different. Indeed, by considering the additional
sparsity assumption (at each stage, at most $s$ decisions incur a nonzero
outcome), we derive optimal regret bounds of different orders. Specifically,
with gains, we obtain an optimal regret guarantee after $T$ stages of order
$\sqrt{T\log s}$, so the classical dependency in the dimension is replaced by
the sparsity size. With losses, we provide matching upper and lower bounds of
order $\sqrt{Ts\log(d)/d}$, which is decreasing in $d$. Eventually, we also
study the bandit setting, and obtain an upper bound of order $\sqrt{Ts\log
(d/s)}$ when outcomes are losses. This bound is proven to be optimal up to the
logarithmic factor $\sqrt{\log(d/s)}$.
|
The Mechanism of Additive Composition | Additive composition (Foltz et al, 1998; Landauer and Dumais, 1997; Mitchell
and Lapata, 2010) is a widely used method for computing meanings of phrases,
which takes the average of vector representations of the constituent words. In
this article, we prove an upper bound for the bias of additive composition,
which is the first theoretical analysis on compositional frameworks from a
machine learning point of view. The bound is written in terms of collocation
strength; we prove that the more exclusively two successive words tend to occur
together, the more accurate one can guarantee their additive composition as an
approximation to the natural phrase vector. Our proof relies on properties of
natural language data that are empirically verified, and can be theoretically
derived from an assumption that the data is generated from a Hierarchical
Pitman-Yor Process. The theory endorses additive composition as a reasonable
operation for calculating meanings of phrases, and suggests ways to improve
additive compositionality, including: transforming entries of distributional
word vectors by a function that meets a specific condition, constructing a
novel type of vector representations to make additive composition sensitive to
word order, and utilizing singular value decomposition to train word vectors.
|
An Introduction to Convolutional Neural Networks | The field of machine learning has taken a dramatic twist in recent times,
with the rise of the Artificial Neural Network (ANN). These biologically
inspired computational models are able to far exceed the performance of
previous forms of artificial intelligence in common machine learning tasks. One
of the most impressive forms of ANN architecture is that of the Convolutional
Neural Network (CNN). CNNs are primarily used to solve difficult image-driven
pattern recognition tasks and with their precise yet simple architecture,
offers a simplified method of getting started with ANNs.
This document provides a brief introduction to CNNs, discussing recently
published papers and newly formed techniques in developing these brilliantly
fantastic image recognition models. This introduction assumes you are familiar
with the fundamentals of ANNs and machine learning.
|
Distributed Machine Learning via Sufficient Factor Broadcasting | Matrix-parametrized models, including multiclass logistic regression and
sparse coding, are used in machine learning (ML) applications ranging from
computer vision to computational biology. When these models are applied to
large-scale ML problems starting at millions of samples and tens of thousands
of classes, their parameter matrix can grow at an unexpected rate, resulting in
high parameter synchronization costs that greatly slow down distributed
learning. To address this issue, we propose a Sufficient Factor Broadcasting
(SFB) computation model for efficient distributed learning of a large family of
matrix-parameterized models, which share the following property: the parameter
update computed on each data sample is a rank-1 matrix, i.e., the outer product
of two "sufficient factors" (SFs). By broadcasting the SFs among worker
machines and reconstructing the update matrices locally at each worker, SFB
improves communication efficiency --- communication costs are linear in the
parameter matrix's dimensions, rather than quadratic --- without affecting
computational correctness. We present a theoretical convergence analysis of
SFB, and empirically corroborate its efficiency on four different
matrix-parametrized ML models.
|
Incremental Truncated LSTD | Balancing between computational efficiency and sample efficiency is an
important goal in reinforcement learning. Temporal difference (TD) learning
algorithms stochastically update the value function, with a linear time
complexity in the number of features, whereas least-squares temporal difference
(LSTD) algorithms are sample efficient but can be quadratic in the number of
features. In this work, we develop an efficient incremental low-rank
LSTD({\lambda}) algorithm that progresses towards the goal of better balancing
computation and sample efficiency. The algorithm reduces the computation and
storage complexity to the number of features times the chosen rank parameter
while summarizing past samples efficiently to nearly obtain the sample
complexity of LSTD. We derive a simulation bound on the solution given by
truncated low-rank approximation, illustrating a bias- variance trade-off
dependent on the choice of rank. We demonstrate that the algorithm effectively
balances computational complexity and sample efficiency for policy evaluation
in a benchmark task and a high-dimensional energy allocation domain.
|
Iterative Instance Segmentation | Existing methods for pixel-wise labelling tasks generally disregard the
underlying structure of labellings, often leading to predictions that are
visually implausible. While incorporating structure into the model should
improve prediction quality, doing so is challenging - manually specifying the
form of structural constraints may be impractical and inference often becomes
intractable even if structural constraints are given. We sidestep this problem
by reducing structured prediction to a sequence of unconstrained prediction
problems and demonstrate that this approach is capable of automatically
discovering priors on shape, contiguity of region predictions and smoothness of
region contours from data without any a priori specification. On the instance
segmentation task, this method outperforms the state-of-the-art, achieving a
mean $\mathrm{AP}^{r}$ of 63.6% at 50% overlap and 43.3% at 70% overlap.
|
Regularized EM Algorithms: A Unified Framework and Statistical
Guarantees | Latent variable models are a fundamental modeling tool in machine learning
applications, but they present significant computational and analytical
challenges. The popular EM algorithm and its variants, is a much used
algorithmic tool; yet our rigorous understanding of its performance is highly
incomplete. Recently, work in Balakrishnan et al. (2014) has demonstrated that
for an important class of problems, EM exhibits linear local convergence. In
the high-dimensional setting, however, the M-step may not be well defined. We
address precisely this setting through a unified treatment using
regularization. While regularization for high-dimensional problems is by now
well understood, the iterative EM algorithm requires a careful balancing of
making progress towards the solution while identifying the right structure
(e.g., sparsity or low-rank). In particular, regularizing the M-step using the
state-of-the-art high-dimensional prescriptions (e.g., Wainwright (2014)) is
not guaranteed to provide this balance. Our algorithm and analysis are linked
in a way that reveals the balance between optimization and statistical errors.
We specialize our general framework to sparse gaussian mixture models,
high-dimensional mixed regression, and regression with missing variables,
obtaining statistical guarantees for each of these examples.
|
Simultaneous Private Learning of Multiple Concepts | We investigate the direct-sum problem in the context of differentially
private PAC learning: What is the sample complexity of solving $k$ learning
tasks simultaneously under differential privacy, and how does this cost compare
to that of solving $k$ learning tasks without privacy? In our setting, an
individual example consists of a domain element $x$ labeled by $k$ unknown
concepts $(c_1,\ldots,c_k)$. The goal of a multi-learner is to output $k$
hypotheses $(h_1,\ldots,h_k)$ that generalize the input examples.
Without concern for privacy, the sample complexity needed to simultaneously
learn $k$ concepts is essentially the same as needed for learning a single
concept. Under differential privacy, the basic strategy of learning each
hypothesis independently yields sample complexity that grows polynomially with
$k$. For some concept classes, we give multi-learners that require fewer
samples than the basic strategy. Unfortunately, however, we also give lower
bounds showing that even for very simple concept classes, the sample cost of
private multi-learning must grow polynomially in $k$.
|
Shaping Proto-Value Functions via Rewards | In this paper, we combine task-dependent reward shaping and task-independent
proto-value functions to obtain reward dependent proto-value functions (RPVFs).
In constructing the RPVFs we are making use of the immediate rewards which are
available during the sampling phase but are not used in the PVF construction.
We show via experiments that learning with an RPVF based representation is
better than learning with just reward shaping or PVFs. In particular, when the
state space is symmetrical and the rewards are asymmetrical, the RPVF capture
the asymmetry better than the PVFs.
|
Algorithms for Differentially Private Multi-Armed Bandits | We present differentially private algorithms for the stochastic Multi-Armed
Bandit (MAB) problem. This is a problem for applications such as adaptive
clinical trials, experiment design, and user-targeted advertising where private
information is connected to individual rewards. Our major contribution is to
show that there exist $(\epsilon, \delta)$ differentially private variants of
Upper Confidence Bound algorithms which have optimal regret, $O(\epsilon^{-1} +
\log T)$. This is a significant improvement over previous results, which only
achieve poly-log regret $O(\epsilon^{-2} \log^{2} T)$, because of our use of a
novel interval-based mechanism. We also substantially improve the bounds of
previous family of algorithms which use a continual release mechanism.
Experiments clearly validate our theoretical bounds.
|
On the convergence of cycle detection for navigational reinforcement
learning | We consider a reinforcement learning framework where agents have to navigate
from start states to goal states. We prove convergence of a cycle-detection
learning algorithm on a class of tasks that we call reducible. Reducible tasks
have an acyclic solution. We also syntactically characterize the form of the
final policy. This characterization can be used to precisely detect the
convergence point in a simulation. Our result demonstrates that even simple
algorithms can be successful in learning a large class of nontrivial tasks. In
addition, our framework is elementary in the sense that we only use basic
concepts to formally prove convergence.
|
Informative Data Projections: A Framework and Two Examples | Methods for Projection Pursuit aim to facilitate the visual exploration of
high-dimensional data by identifying interesting low-dimensional projections. A
major challenge is the design of a suitable quality metric of projections,
commonly referred to as the projection index, to be maximized by the Projection
Pursuit algorithm. In this paper, we introduce a new information-theoretic
strategy for tackling this problem, based on quantifying the amount of
information the projection conveys to a user given their prior beliefs about
the data. The resulting projection index is a subjective quantity, explicitly
dependent on the intended user. As a useful illustration, we developed this
idea for two particular kinds of prior beliefs. The first kind leads to PCA
(Principal Component Analysis), shining new light on when PCA is (not)
appropriate. The second kind leads to a novel projection index, the
maximization of which can be regarded as a robust variant of PCA. We show how
this projection index, though non-convex, can be effectively maximized using a
modified power method as well as using a semidefinite programming relaxation.
The usefulness of this new projection index is demonstrated in comparative
empirical experiments against PCA and a popular Projection Pursuit method.
|
Multiagent Cooperation and Competition with Deep Reinforcement Learning | Multiagent systems appear in most social, economical, and political
situations. In the present work we extend the Deep Q-Learning Network
architecture proposed by Google DeepMind to multiagent environments and
investigate how two agents controlled by independent Deep Q-Networks interact
in the classic videogame Pong. By manipulating the classical rewarding scheme
of Pong we demonstrate how competitive and collaborative behaviors emerge.
Competitive agents learn to play and score efficiently. Agents trained under
collaborative rewarding schemes find an optimal strategy to keep the ball in
the game as long as possible. We also describe the progression from competitive
to collaborative behavior. The present work demonstrates that Deep Q-Networks
can become a practical tool for studying the decentralized learning of
multiagent systems living in highly complex environments.
|
Efficient Sum of Outer Products Dictionary Learning (SOUP-DIL) - The
$\ell_0$ Method | The sparsity of natural signals and images in a transform domain or
dictionary has been extensively exploited in several applications such as
compression, denoising and inverse problems. More recently, data-driven
adaptation of synthesis dictionaries has shown promise in many applications
compared to fixed or analytical dictionary models. However, dictionary learning
problems are typically non-convex and NP-hard, and the usual alternating
minimization approaches for these problems are often computationally expensive,
with the computations dominated by the NP-hard synthesis sparse coding step. In
this work, we investigate an efficient method for $\ell_{0}$ "norm"-based
dictionary learning by first approximating the training data set with a sum of
sparse rank-one matrices and then using a block coordinate descent approach to
estimate the unknowns. The proposed block coordinate descent algorithm involves
efficient closed-form solutions. In particular, the sparse coding step involves
a simple form of thresholding. We provide a convergence analysis for the
proposed block coordinate descent approach. Our numerical experiments show the
promising performance and significant speed-ups provided by our method over the
classical K-SVD scheme in sparse signal representation and image denoising.
|
Designing High-Fidelity Single-Shot Three-Qubit Gates: A Machine
Learning Approach | Three-qubit quantum gates are key ingredients for quantum error correction
and quantum information processing. We generate quantum-control procedures to
design three types of three-qubit gates, namely Toffoli, Controlled-Not-Not and
Fredkin gates. The design procedures are applicable to a system comprising
three nearest-neighbor-coupled superconducting artificial atoms. For each
three-qubit gate, the numerical simulation of the proposed scheme achieves
99.9% fidelity, which is an accepted threshold fidelity for fault-tolerant
quantum computing. We test our procedure in the presence of decoherence-induced
noise as well as show its robustness against random external noise generated by
the control electronics. The three-qubit gates are designed via the machine
learning algorithm called Subspace-Selective Self-Adaptive Differential
Evolution (SuSSADE).
|
MidRank: Learning to rank based on subsequences | We present a supervised learning to rank algorithm that effectively orders
images by exploiting the structure in image sequences. Most often in the
supervised learning to rank literature, ranking is approached either by
analyzing pairs of images or by optimizing a list-wise surrogate loss function
on full sequences. In this work we propose MidRank, which learns from
moderately sized sub-sequences instead. These sub-sequences contain useful
structural ranking information that leads to better learnability during
training and better generalization during testing. By exploiting sub-sequences,
the proposed MidRank improves ranking accuracy considerably on an extensive
array of image ranking applications and datasets.
|
Learning Directed Acyclic Graphs with Penalized Neighbourhood Regression | We study a family of regularized score-based estimators for learning the
structure of a directed acyclic graph (DAG) for a multivariate normal
distribution from high-dimensional data with $p\gg n$. Our main results
establish support recovery guarantees and deviation bounds for a family of
penalized least-squares estimators under concave regularization without
assuming prior knowledge of a variable ordering. These results apply to a
variety of practical situations that allow for arbitrary nondegenerate
covariance structures as well as many popular regularizers including the MCP,
SCAD, $\ell_{0}$ and $\ell_{1}$. The proof relies on interpreting a DAG as a
recursive linear structural equation model, which reduces the estimation
problem to a series of neighbourhood regressions. We provide a novel
statistical analysis of these neighbourhood problems, establishing uniform
control over the superexponential family of neighbourhoods associated with a
Gaussian distribution. We then apply these results to study the statistical
properties of score-based DAG estimators, learning causal DAGs, and inferring
conditional independence relations via graphical models. Our results
yield---for the first time---finite-sample guarantees for structure learning of
Gaussian DAGs in high-dimensions via score-based estimation.
|
Robotic Search & Rescue via Online Multi-task Reinforcement Learning | Reinforcement learning (RL) is a general and well-known method that a robot
can use to learn an optimal control policy to solve a particular task. We would
like to build a versatile robot that can learn multiple tasks, but using RL for
each of them would be prohibitively expensive in terms of both time and
wear-and-tear on the robot. To remedy this problem, we use the Policy Gradient
Efficient Lifelong Learning Algorithm (PG-ELLA), an online multi-task RL
algorithm that enables the robot to efficiently learn multiple consecutive
tasks by sharing knowledge between these tasks to accelerate learning and
improve performance. We implemented and evaluated three RL methods--Q-learning,
policy gradient RL, and PG-ELLA--on a ground robot whose task is to find a
target object in an environment under different surface conditions. In this
paper, we discuss our implementations as well as present an empirical analysis
of their learning performance.
|
How do the naive Bayes classifier and the Support Vector Machine compare
in their ability to forecast the Stock Exchange of Thailand? | This essay investigates the question of how the naive Bayes classifier and
the support vector machine compare in their ability to forecast the Stock
Exchange of Thailand. The theory behind the SVM and the naive Bayes classifier
is explored. The algorithms are trained using data from the month of January
2010, extracted from the MarketWatch.com website. Input features are selected
based on previous studies of the SET100 Index. The Weka 3 software is used to
create models from the labeled training data. Mean squared error and proportion
of correctly classified instances, and a number of other error measurements are
the used to compare the two algorithms. This essay shows that these two
algorithms are currently not advanced enough to accurately model the stock
exchange. Nevertheless, the naive Bayes is better than the support vector
machine at predicting the Stock Exchange of Thailand.
|
Multiple-Instance Learning: Radon-Nikodym Approach to Distribution
Regression Problem | For distribution regression problem, where a bag of $x$--observations is
mapped to a single $y$ value, a one--step solution is proposed. The problem of
random distribution to random value is transformed to random vector to random
value by taking distribution moments of $x$ observations in a bag as random
vector. Then Radon--Nikodym or least squares theory can be applied, what give
$y(x)$ estimator. The probability distribution of $y$ is also obtained, what
requires solving generalized eigenvalues problem, matrix spectrum (not
depending on $x$) give possible $y$ outcomes and depending on $x$ probabilities
of outcomes can be obtained by projecting the distribution with fixed $x$ value
(delta--function) to corresponding eigenvector. A library providing numerically
stable polynomial basis for these calculations is available, what make the
proposed approach practical.
|
Position paper: a general framework for applying machine learning
techniques in operating room | In this position paper we describe a general framework for applying machine
learning and pattern recognition techniques in healthcare. In particular, we
are interested in providing an automated tool for monitoring and incrementing
the level of awareness in the operating room and for identifying human errors
which occur during the laparoscopy surgical operation. The framework that we
present is divided in three different layers: each layer implements algorithms
which have an increasing level of complexity and which perform functionality
with an higher degree of abstraction. In the first layer, raw data collected
from sensors in the operating room during surgical operation, they are
pre-processed and aggregated. The results of this initial phase are transferred
to a second layer, which implements pattern recognition techniques and extract
relevant features from the data. Finally, in the last layer, expert systems are
employed to take high level decisions, which represent the final output of the
system.
|
A Short Survey on Data Clustering Algorithms | With rapidly increasing data, clustering algorithms are important tools for
data analytics in modern research. They have been successfully applied to a
wide range of domains; for instance, bioinformatics, speech recognition, and
financial analysis. Formally speaking, given a set of data instances, a
clustering algorithm is expected to divide the set of data instances into the
subsets which maximize the intra-subset similarity and inter-subset
dissimilarity, where a similarity measure is defined beforehand. In this work,
the state-of-the-arts clustering algorithms are reviewed from design concept to
methodology; Different clustering paradigms are discussed. Advanced clustering
algorithms are also discussed. After that, the existing clustering evaluation
metrics are reviewed. A summary with future insights is provided at the end.
|
Aspect-based Opinion Summarization with Convolutional Neural Networks | This paper considers Aspect-based Opinion Summarization (AOS) of reviews on
particular products. To enable real applications, an AOS system needs to
address two core subtasks, aspect extraction and sentiment classification. Most
existing approaches to aspect extraction, which use linguistic analysis or
topic modeling, are general across different products but not precise enough or
suitable for particular products. Instead we take a less general but more
precise scheme, directly mapping each review sentence into pre-defined aspects.
To tackle aspect mapping and sentiment classification, we propose two
Convolutional Neural Network (CNN) based methods, cascaded CNN and multitask
CNN. Cascaded CNN contains two levels of convolutional networks. Multiple CNNs
at level 1 deal with aspect mapping task, and a single CNN at level 2 deals
with sentiment classification. Multitask CNN also contains multiple aspect CNNs
and a sentiment CNN, but different networks share the same word embeddings.
Experimental results indicate that both cascaded and multitask CNNs outperform
SVM-based methods by large margins. Multitask CNN generally performs better
than cascaded CNN.
|
Proximal gradient method for huberized support vector machine | The Support Vector Machine (SVM) has been used in a wide variety of
classification problems. The original SVM uses the hinge loss function, which
is non-differentiable and makes the problem difficult to solve in particular
for regularized SVMs, such as with $\ell_1$-regularization. This paper
considers the Huberized SVM (HSVM), which uses a differentiable approximation
of the hinge loss function. We first explore the use of the Proximal Gradient
(PG) method to solving binary-class HSVM (B-HSVM) and then generalize it to
multi-class HSVM (M-HSVM). Under strong convexity assumptions, we show that our
algorithm converges linearly. In addition, we give a finite convergence result
about the support of the solution, based on which we further accelerate the
algorithm by a two-stage method. We present extensive numerical experiments on
both synthetic and real datasets which demonstrate the superiority of our
methods over some state-of-the-art methods for both binary- and multi-class
SVMs.
|
Asynchronous adaptive networks | In a recent article [1] we surveyed advances related to adaptation, learning,
and optimization over synchronous networks. Various distributed strategies were
discussed that enable a collection of networked agents to interact locally in
response to streaming data and to continually learn and adapt to track drifts
in the data and models. Under reasonable technical conditions on the data, the
adaptive networks were shown to be mean-square stable in the slow adaptation
regime, and their mean-square-error performance and convergence rate were
characterized in terms of the network topology and data statistical moments
[2]. Classical results for single-agent adaptation and learning were recovered
as special cases. Following the works [3]-[5], this chapter complements the
exposition from [1] and extends the results to asynchronous networks. The
operation of this class of networks can be subject to various sources of
uncertainties that influence their dynamic behavior, including randomly
changing topologies, random link failures, random data arrival times, and
agents turning on and off randomly. In an asynchronous environment, agents may
stop updating their solutions or may stop sending or receiving information in a
random manner and without coordination with other agents. The presentation will
reveal that the mean-square-error performance of asynchronous networks remains
largely unaltered compared to synchronous networks. The results justify the
remarkable resilience of cooperative networks in the face of random events.
|
Non-adaptive Group Testing on Graphs | Grebinski and Kucherov (1998) and Alon et al. (2004-2005) study the problem
of learning a hidden graph for some especial cases, such as hamiltonian cycle,
cliques, stars, and matchings. This problem is motivated by problems in
chemical reactions, molecular biology and genome sequencing.
In this paper, we present a generalization of this problem. Precisely, we
consider a graph G and a subgraph H of G and we assume that G contains exactly
one defective subgraph isomorphic to H. The goal is to find the defective
subgraph by testing whether an induced subgraph contains an edge of the
defective subgraph, with the minimum number of tests. We present an upper bound
for the number of tests to find the defective subgraph by using the symmetric
and high probability variation of Lov\'asz Local Lemma.
|
On Learning to Think: Algorithmic Information Theory for Novel
Combinations of Reinforcement Learning Controllers and Recurrent Neural World
Models | This paper addresses the general problem of reinforcement learning (RL) in
partially observable environments. In 2013, our large RL recurrent neural
networks (RNNs) learned from scratch to drive simulated cars from
high-dimensional video input. However, real brains are more powerful in many
ways. In particular, they learn a predictive model of their initially unknown
environment, and somehow use it for abstract (e.g., hierarchical) planning and
reasoning. Guided by algorithmic information theory, we describe RNN-based AIs
(RNNAIs) designed to do the same. Such an RNNAI can be trained on never-ending
sequences of tasks, some of them provided by the user, others invented by the
RNNAI itself in a curious, playful fashion, to improve its RNN-based world
model. Unlike our previous model-building RNN-based RL machines dating back to
1990, the RNNAI learns to actively query its model for abstract reasoning and
planning and decision making, essentially "learning to think." The basic ideas
of this report can be applied to many other cases where one RNN-like system
exploits the algorithmic information content of another. They are taken from a
grant proposal submitted in Fall 2014, and also explain concepts such as
"mirror neurons." Experimental results will be described in separate papers.
|
Scalable and Accurate Online Feature Selection for Big Data | Feature selection is important in many big data applications. Two critical
challenges closely associate with big data. Firstly, in many big data
applications, the dimensionality is extremely high, in millions, and keeps
growing. Secondly, big data applications call for highly scalable feature
selection algorithms in an online manner such that each feature can be
processed in a sequential scan. We present SAOLA, a Scalable and Accurate
OnLine Approach for feature selection in this paper. With a theoretical
analysis on bounds of the pairwise correlations between features, SAOLA employs
novel pairwise comparison techniques and maintain a parsimonious model over
time in an online manner. Furthermore, to deal with upcoming features that
arrive by groups, we extend the SAOLA algorithm, and then propose a new
group-SAOLA algorithm for online group feature selection. The group-SAOLA
algorithm can online maintain a set of feature groups that is sparse at the
levels of both groups and individual features simultaneously. An empirical
study using a series of benchmark real data sets shows that our two algorithms,
SAOLA and group-SAOLA, are scalable on data sets of extremely high
dimensionality, and have superior performance over the state-of-the-art feature
selection methods.
|
Cost-aware Pre-training for Multiclass Cost-sensitive Deep Learning | Deep learning has been one of the most prominent machine learning techniques
nowadays, being the state-of-the-art on a broad range of applications where
automatic feature extraction is needed. Many such applications also demand
varying costs for different types of mis-classification errors, but it is not
clear whether or how such cost information can be incorporated into deep
learning to improve performance. In this work, we propose a novel cost-aware
algorithm that takes into account the cost information into not only the
training stage but also the pre-training stage of deep learning. The approach
allows deep learning to conduct automatic feature extraction with the cost
information effectively. Extensive experimental results demonstrate that the
proposed approach outperforms other deep learning models that do not digest the
cost information in the pre-training stage.
|
k-Nearest Neighbour Classification of Datasets with a Family of
Distances | The $k$-nearest neighbour ($k$-NN) classifier is one of the oldest and most
important supervised learning algorithms for classifying datasets.
Traditionally the Euclidean norm is used as the distance for the $k$-NN
classifier. In this thesis we investigate the use of alternative distances for
the $k$-NN classifier.
We start by introducing some background notions in statistical machine
learning. We define the $k$-NN classifier and discuss Stone's theorem and the
proof that $k$-NN is universally consistent on the normed space $R^d$. We then
prove that $k$-NN is universally consistent if we take a sequence of random
norms (that are independent of the sample and the query) from a family of norms
that satisfies a particular boundedness condition. We extend this result by
replacing norms with distances based on uniformly locally Lipschitz functions
that satisfy certain conditions. We discuss the limitations of Stone's lemma
and Stone's theorem, particularly with respect to quasinorms and adaptively
choosing a distance for $k$-NN based on the labelled sample. We show the
universal consistency of a two stage $k$-NN type classifier where we select the
distance adaptively based on a split labelled sample and the query. We conclude
by giving some examples of improvements of the accuracy of classifying various
datasets using the above techniques.
|
Decoding Hidden Markov Models Faster Than Viterbi Via Online
Matrix-Vector (max, +)-Multiplication | In this paper, we present a novel algorithm for the maximum a posteriori
decoding (MAPD) of time-homogeneous Hidden Markov Models (HMM), improving the
worst-case running time of the classical Viterbi algorithm by a logarithmic
factor. In our approach, we interpret the Viterbi algorithm as a repeated
computation of matrix-vector $(\max, +)$-multiplications. On time-homogeneous
HMMs, this computation is online: a matrix, known in advance, has to be
multiplied with several vectors revealed one at a time. Our main contribution
is an algorithm solving this version of matrix-vector $(\max,+)$-multiplication
in subquadratic time, by performing a polynomial preprocessing of the matrix.
Employing this fast multiplication algorithm, we solve the MAPD problem in
$O(mn^2/ \log n)$ time for any time-homogeneous HMM of size $n$ and observation
sequence of length $m$, with an extra polynomial preprocessing cost negligible
for $m > n$. To the best of our knowledge, this is the first algorithm for the
MAPD problem requiring subquadratic time per observation, under the only
assumption -- usually verified in practice -- that the transition probability
matrix does not change with time.
|
Learning Using 1-Local Membership Queries | Classic machine learning algorithms learn from labelled examples. For
example, to design a machine translation system, a typical training set will
consist of English sentences and their translation. There is a stronger model,
in which the algorithm can also query for labels of new examples it creates.
E.g, in the translation task, the algorithm can create a new English sentence,
and request its translation from the user during training. This combination of
examples and queries has been widely studied. Yet, despite many theoretical
results, query algorithms are almost never used. One of the main causes for
this is a report (Baum and Lang, 1992) on very disappointing empirical
performance of a query algorithm. These poor results were mainly attributed to
the fact that the algorithm queried for labels of examples that are artificial,
and impossible to interpret by humans.
In this work we study a new model of local membership queries (Awasthi et
al., 2012), which tries to resolve the problem of artificial queries. In this
model, the algorithm is only allowed to query the labels of examples which are
close to examples from the training set. E.g., in translation, the algorithm
can change individual words in a sentence it has already seen, and then ask for
the translation. In this model, the examples queried by the algorithm will be
close to natural examples and hence, hopefully, will not appear as artificial
or random. We focus on 1-local queries (i.e., queries of distance 1 from an
example in the training sample). We show that 1-local membership queries are
already stronger than the standard learning model. We also present an
experiment on a well known NLP task of sentiment analysis. In this experiment,
the users were asked to provide more information than merely indicating the
label. We present results that illustrate that this extra information is
beneficial in practice.
|
MOCICE-BCubed F$_1$: A New Evaluation Measure for Biclustering
Algorithms | The validation of biclustering algorithms remains a challenging task, even
though a number of measures have been proposed for evaluating the quality of
these algorithms. Although no criterion is universally accepted as the overall
best, a number of meta-evaluation conditions to be satisfied by biclustering
algorithms have been enunciated. In this work, we present MOCICE-BCubed F$_1$,
a new external measure for evaluating biclusterings, in the scenario where gold
standard annotations are available for both the object clusters and the
associated feature subspaces. Our proposal relies on the so-called
micro-objects transformation and satisfies the most comprehensive set of
meta-evaluation conditions so far enunciated for biclusterings. Additionally,
the proposed measure adequately handles the occurrence of overlapping in both
the object and feature spaces. Moreover, when used for evaluating traditional
clusterings, which are viewed as a particular case of biclustering, the
proposed measure also satisfies the most comprehensive set of meta-evaluation
conditions so far enunciated for this task.
|
Towards Dropout Training for Convolutional Neural Networks | Recently, dropout has seen increasing use in deep learning. For deep
convolutional neural networks, dropout is known to work well in fully-connected
layers. However, its effect in convolutional and pooling layers is still not
clear. This paper demonstrates that max-pooling dropout is equivalent to
randomly picking activation based on a multinomial distribution at training
time. In light of this insight, we advocate employing our proposed
probabilistic weighted pooling, instead of commonly used max-pooling, to act as
model averaging at test time. Empirical evidence validates the superiority of
probabilistic weighted pooling. We also empirically show that the effect of
convolutional dropout is not trivial, despite the dramatically reduced
possibility of over-fitting due to the convolutional architecture. Elaborately
designing dropout training simultaneously in max-pooling and fully-connected
layers, we achieve state-of-the-art performance on MNIST, and very competitive
results on CIFAR-10 and CIFAR-100, relative to other approaches without data
augmentation. Finally, we compare max-pooling dropout and stochastic pooling,
both of which introduce stochasticity based on multinomial distributions at
pooling stage.
|
Sequential visibility-graph motifs | Visibility algorithms transform time series into graphs and encode dynamical
information in their topology, paving the way for graph-theoretical time series
analysis as well as building a bridge between nonlinear dynamics and network
science. In this work we introduce and study the concept of sequential
visibility graph motifs, smaller substructures of n consecutive nodes that
appear with characteristic frequencies. We develop a theory to compute in an
exact way the motif profiles associated to general classes of deterministic and
stochastic dynamics. We find that this simple property is indeed a highly
informative and computationally efficient feature capable to distinguish among
different dynamics and robust against noise contamination. We finally confirm
that it can be used in practice to perform unsupervised learning, by extracting
motif profiles from experimental heart-rate series and being able, accordingly,
to disentangle meditative from other relaxation states. Applications of this
general theory include the automatic classification and description of
physical, biological, and financial time series.
|
Taxonomy grounded aggregation of classifiers with different label sets | We describe the problem of aggregating the label predictions of diverse
classifiers using a class taxonomy. Such a taxonomy may not have been available
or referenced when the individual classifiers were designed and trained, yet
mapping the output labels into the taxonomy is desirable to integrate the
effort spent in training the constituent classifiers. A hierarchical taxonomy
representing some domain knowledge may be different from, but partially
mappable to, the label sets of the individual classifiers. We present a
heuristic approach and a principled graphical model to aggregate the label
predictions by grounding them into the available taxonomy. Our model aggregates
the labels using the taxonomy structure as constraints to find the most likely
hierarchically consistent class. We experimentally validate our proposed method
on image and text classification tasks.
|
Reinforcement Learning Applied to an Electric Water Heater: From Theory
to Practice | Electric water heaters have the ability to store energy in their water buffer
without impacting the comfort of the end user. This feature makes them a prime
candidate for residential demand response. However, the stochastic and
nonlinear dynamics of electric water heaters, makes it challenging to harness
their flexibility. Driven by this challenge, this paper formulates the
underlying sequential decision-making problem as a Markov decision process and
uses techniques from reinforcement learning. Specifically, we apply an
auto-encoder network to find a compact feature representation of the sensor
measurements, which helps to mitigate the curse of dimensionality. A wellknown
batch reinforcement learning technique, fitted Q-iteration, is used to find a
control policy, given this feature representation. In a simulation-based
experiment using an electric water heater with 50 temperature sensors, the
proposed method was able to achieve good policies much faster than when using
the full state information. In a lab experiment, we apply fitted Q-iteration to
an electric water heater with eight temperature sensors. Further reducing the
state vector did not improve the results of fitted Q-iteration. The results of
the lab experiment, spanning 40 days, indicate that compared to a thermostat
controller, the presented approach was able to reduce the total cost of energy
consumption of the electric water heater by 15%.
|
Fast k-Nearest Neighbour Search via Dynamic Continuous Indexing | Existing methods for retrieving k-nearest neighbours suffer from the curse of
dimensionality. We argue this is caused in part by inherent deficiencies of
space partitioning, which is the underlying strategy used by most existing
methods. We devise a new strategy that avoids partitioning the vector space and
present a novel randomized algorithm that runs in time linear in dimensionality
of the space and sub-linear in the intrinsic dimensionality and the size of the
dataset and takes space constant in dimensionality of the space and linear in
the size of the dataset. The proposed algorithm allows fine-grained control
over accuracy and speed on a per-query basis, automatically adapts to
variations in data density, supports dynamic updates to the dataset and is
easy-to-implement. We show appealing theoretical properties and demonstrate
empirically that the proposed algorithm outperforms locality-sensitivity
hashing (LSH) in terms of approximation quality, speed and space efficiency.
|
Loss Functions for Top-k Error: Analysis and Insights | In order to push the performance on realistic computer vision tasks, the
number of classes in modern benchmark datasets has significantly increased in
recent years. This increase in the number of classes comes along with increased
ambiguity between the class labels, raising the question if top-1 error is the
right performance measure. In this paper, we provide an extensive comparison
and evaluation of established multiclass methods comparing their top-k
performance both from a practical as well as from a theoretical perspective.
Moreover, we introduce novel top-k loss functions as modifications of the
softmax and the multiclass SVM losses and provide efficient optimization
schemes for them. In the experiments, we compare on various datasets all of the
proposed and established methods for top-k error optimization. An interesting
insight of this paper is that the softmax loss yields competitive top-k
performance for all k simultaneously. For a specific top-k error, our new top-k
losses lead typically to further improvements while being faster to train than
the softmax.
|
Attribute2Image: Conditional Image Generation from Visual Attributes | This paper investigates a novel problem of generating images from visual
attributes. We model the image as a composite of foreground and background and
develop a layered generative model with disentangled latent variables that can
be learned end-to-end using a variational auto-encoder. We experiment with
natural images of faces and birds and demonstrate that the proposed models are
capable of generating realistic and diverse samples with disentangled latent
representations. We use a general energy minimization algorithm for posterior
inference of latent variables given novel images. Therefore, the learned
generative models show excellent quantitative and visual results in the tasks
of attribute-conditioned image reconstruction and completion.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.