title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Object-based World Modeling in Semi-Static Environments with Dependent
Dirichlet-Process Mixtures | To accomplish tasks in human-centric indoor environments, robots need to
represent and understand the world in terms of objects and their attributes. We
refer to this attribute-based representation as a world model, and consider how
to acquire it via noisy perception and maintain it over time, as objects are
added, changed, and removed in the world. Previous work has framed this as
multiple-target tracking problem, where objects are potentially in motion at
all times. Although this approach is general, it is computationally expensive.
We argue that such generality is not needed in typical world modeling tasks,
where objects only change state occasionally. More efficient approaches are
enabled by restricting ourselves to such semi-static environments.
We consider a previously-proposed clustering-based world modeling approach
that assumed static environments, and extend it to semi-static domains by
applying a dependent Dirichlet-process (DDP) mixture model. We derive a novel
MAP inference algorithm under this model, subject to data association
constraints. We demonstrate our approach improves computational performance in
semi-static environments.
|
Centroid Based Binary Tree Structured SVM for Multi Classification | Support Vector Machines (SVMs) were primarily designed for 2-class
classification. But they have been extended for N-class classification also
based on the requirement of multiclasses in the practical applications.
Although N-class classification using SVM has considerable research attention,
getting minimum number of classifiers at the time of training and testing is
still a continuing research. We propose a new algorithm CBTS-SVM (Centroid
based Binary Tree Structured SVM) which addresses this issue. In this we build
a binary tree of SVM models based on the similarity of the class labels by
finding their distance from the corresponding centroids at the root level. The
experimental results demonstrates the comparable accuracy for CBTS with OVO
with reasonable gamma and cost values. On the other hand when CBTS is compared
with OVA, it gives the better accuracy with reduced training time and testing
time. Furthermore CBTS is also scalable as it is able to handle the large data
sets.
|
Recognizing Semantic Features in Faces using Deep Learning | The human face constantly conveys information, both consciously and
subconsciously. However, as basic as it is for humans to visually interpret
this information, it is quite a big challenge for machines. Conventional
semantic facial feature recognition and analysis techniques are already in use
and are based on physiological heuristics, but they suffer from lack of
robustness and high computation time. This thesis aims to explore ways for
machines to learn to interpret semantic information available in faces in an
automated manner without requiring manual design of feature detectors, using
the approach of Deep Learning. This thesis provides a study of the effects of
various factors and hyper-parameters of deep neural networks in the process of
determining an optimal network configuration for the task of semantic facial
feature recognition. This thesis explores the effectiveness of the system to
recognize the various semantic features (like emotions, age, gender, ethnicity
etc.) present in faces. Furthermore, the relation between the effect of
high-level concepts on low level features is explored through an analysis of
the similarities in low-level descriptors of different semantic features. This
thesis also demonstrates a novel idea of using a deep network to generate 3-D
Active Appearance Models of faces from real-world 2-D images.
For a more detailed report on this work, please see [arXiv:1512.00743v1].
|
Zero-Shot Event Detection by Multimodal Distributional Semantic
Embedding of Videos | We propose a new zero-shot Event Detection method by Multi-modal
Distributional Semantic embedding of videos. Our model embeds object and action
concepts as well as other available modalities from videos into a
distributional semantic space. To our knowledge, this is the first Zero-Shot
event detection model that is built on top of distributional semantics and
extends it in the following directions: (a) semantic embedding of multimodal
information in videos (with focus on the visual modalities), (b) automatically
determining relevance of concepts/attributes to a free text query, which could
be useful for other applications, and (c) retrieving videos by free text event
query (e.g., "changing a vehicle tire") based on their content. We embed videos
into a distributional semantic space and then measure the similarity between
videos and the event query in a free text form. We validated our method on the
large TRECVID MED (Multimedia Event Detection) challenge. Using only the event
title as a query, our method outperformed the state-of-the-art that uses big
descriptions from 12.6% to 13.5% with MAP metric and 0.73 to 0.83 with ROC-AUC
metric. It is also an order of magnitude faster.
|
Protein secondary structure prediction using deep convolutional neural
fields | Protein secondary structure (SS) prediction is important for studying protein
structure and function. When only the sequence (profile) information is used as
input feature, currently the best predictors can obtain ~80% Q3 accuracy, which
has not been improved in the past decade. Here we present DeepCNF (Deep
Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep
Learning extension of Conditional Neural Fields (CNF), which is an integration
of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can
model not only complex sequence-structure relationship by a deep hierarchical
architecture, but also interdependency between adjacent SS labels, so it is
much more powerful than CNF. Experimental results show that DeepCNF can obtain
~84% Q3 accuracy, ~85% SOV score, and ~72% Q8 accuracy, respectively, on the
CASP and CAMEO test proteins, greatly outperforming currently popular
predictors. As a general framework, DeepCNF can be used to predict other
protein structure properties such as contact number, disorder regions, and
solvent accessibility.
|
Innovation Pursuit: A New Approach to Subspace Clustering | In subspace clustering, a group of data points belonging to a union of
subspaces are assigned membership to their respective subspaces. This paper
presents a new approach dubbed Innovation Pursuit (iPursuit) to the problem of
subspace clustering using a new geometrical idea whereby subspaces are
identified based on their relative novelties. We present two frameworks in
which the idea of innovation pursuit is used to distinguish the subspaces.
Underlying the first framework is an iterative method that finds the subspaces
consecutively by solving a series of simple linear optimization problems, each
searching for a direction of innovation in the span of the data potentially
orthogonal to all subspaces except for the one to be identified in one step of
the algorithm. A detailed mathematical analysis is provided establishing
sufficient conditions for iPursuit to correctly cluster the data. The proposed
approach can provably yield exact clustering even when the subspaces have
significant intersections. It is shown that the complexity of the iterative
approach scales only linearly in the number of data points and subspaces, and
quadratically in the dimension of the subspaces. The second framework
integrates iPursuit with spectral clustering to yield a new variant of
spectral-clustering-based algorithms. The numerical simulations with both real
and synthetic data demonstrate that iPursuit can often outperform the
state-of-the-art subspace clustering algorithms, more so for subspaces with
significant intersections, and that it significantly improves the
state-of-the-art result for subspace-segmentation-based face clustering.
|
Neural Enquirer: Learning to Query Tables with Natural Language | We proposed Neural Enquirer as a neural network architecture to execute a
natural language (NL) query on a knowledge-base (KB) for answers. Basically,
Neural Enquirer finds the distributed representation of a query and then
executes it on knowledge-base tables to obtain the answer as one of the values
in the tables. Unlike similar efforts in end-to-end training of semantic
parsers, Neural Enquirer is fully "neuralized": it not only gives
distributional representation of the query and the knowledge-base, but also
realizes the execution of compositional queries as a series of differentiable
operations, with intermediate results (consisting of annotations of the tables
at different levels) saved on multiple layers of memory. Neural Enquirer can be
trained with gradient descent, with which not only the parameters of the
controlling components and semantic parsing component, but also the embeddings
of the tables and query words can be learned from scratch. The training can be
done in an end-to-end fashion, but it can take stronger guidance, e.g., the
step-by-step supervision for complicated queries, and benefit from it. Neural
Enquirer is one step towards building neural network systems which seek to
understand language by executing it on real-world. Our experiments show that
Neural Enquirer can learn to execute fairly complicated NL queries on tables
with rich structures.
|
Fast Low-Rank Matrix Learning with Nonconvex Regularization | Low-rank modeling has a lot of important applications in machine learning,
computer vision and social network analysis. While the matrix rank is often
approximated by the convex nuclear norm, the use of nonconvex low-rank
regularizers has demonstrated better recovery performance. However, the
resultant optimization problem is much more challenging. A very recent
state-of-the-art is based on the proximal gradient algorithm. However, it
requires an expensive full SVD in each proximal step. In this paper, we show
that for many commonly-used nonconvex low-rank regularizers, a cutoff can be
derived to automatically threshold the singular values obtained from the
proximal operator. This allows the use of power method to approximate the SVD
efficiently. Besides, the proximal operator can be reduced to that of a much
smaller matrix projected onto this leading subspace. Convergence, with a rate
of O(1/T) where T is the number of iterations, can be guaranteed. Extensive
experiments are performed on matrix completion and robust principal component
analysis. The proposed method achieves significant speedup over the
state-of-the-art. Moreover, the matrix solution obtained is more accurate and
has a lower rank than that of the traditional nuclear norm regularizer.
|
Bag Reference Vector for Multi-instance Learning | Multi-instance learning (MIL) has a wide range of applications due to its
distinctive characteristics. Although many state-of-the-art algorithms have
achieved decent performances, a plurality of existing methods solve the problem
only in instance level rather than excavating relations among bags. In this
paper, we propose an efficient algorithm to describe each bag by a
corresponding feature vector via comparing it with other bags. In other words,
the crucial information of a bag is extracted from the similarity between that
bag and other reference bags. In addition, we apply extensions of Hausdorff
distance to representing the similarity, to a certain extent, overcoming the
key challenge of MIL problem, the ambiguity of instances' labels in positive
bags. Experimental results on benchmarks and text categorization tasks show
that the proposed method outperforms the previous state-of-the-art by a large
margin.
|
Bayesian Matrix Completion via Adaptive Relaxed Spectral Regularization | Bayesian matrix completion has been studied based on a low-rank matrix
factorization formulation with promising results. However, little work has been
done on Bayesian matrix completion based on the more direct spectral
regularization formulation. We fill this gap by presenting a novel Bayesian
matrix completion method based on spectral regularization. In order to
circumvent the difficulties of dealing with the orthonormality constraints of
singular vectors, we derive a new equivalent form with relaxed constraints,
which then leads us to design an adaptive version of spectral regularization
feasible for Bayesian inference. Our Bayesian method requires no parameter
tuning and can infer the number of latent factors automatically. Experiments on
synthetic and real datasets demonstrate encouraging results on rank recovery
and collaborative filtering, with notably good results for very sparse
matrices.
|
Deep Reinforcement Learning with Attention for Slate Markov Decision
Processes with High-Dimensional States and Actions | Many real-world problems come with action spaces represented as feature
vectors. Although high-dimensional control is a largely unsolved problem, there
has recently been progress for modest dimensionalities. Here we report on a
successful attempt at addressing problems of dimensionality as high as $2000$,
of a particular form. Motivated by important applications such as
recommendation systems that do not fit the standard reinforcement learning
frameworks, we introduce Slate Markov Decision Processes (slate-MDPs). A
Slate-MDP is an MDP with a combinatorial action space consisting of slates
(tuples) of primitive actions of which one is executed in an underlying MDP.
The agent does not control the choice of this executed action and the action
might not even be from the slate, e.g., for recommendation systems for which
all recommendations can be ignored. We use deep Q-learning based on feature
representations of both the state and action to learn the value of whole
slates. Unlike existing methods, we optimize for both the combinatorial and
sequential aspects of our tasks. The new agent's superiority over agents that
either ignore the combinatorial or sequential long-term value aspect is
demonstrated on a range of environments with dynamics from a real-world
recommendation system. Further, we use deep deterministic policy gradients to
learn a policy that for each position of the slate, guides attention towards
the part of the action space in which the value is the highest and we only
evaluate actions in this area. The attention is used within a sequentially
greedy procedure leveraging submodularity. Finally, we show how introducing
risk-seeking can dramatically improve the agents performance and ability to
discover more far reaching strategies.
|
Building Memory with Concept Learning Capabilities from Large-scale
Knowledge Base | We present a new perspective on neural knowledge base (KB) embeddings, from
which we build a framework that can model symbolic knowledge in the KB together
with its learning process. We show that this framework well regularizes
previous neural KB embedding model for superior performance in reasoning tasks,
while having the capabilities of dealing with unseen entities, that is, to
learn their embeddings from natural language descriptions, which is very like
human's behavior of learning semantic concepts.
|
MXNet: A Flexible and Efficient Machine Learning Library for
Heterogeneous Distributed Systems | MXNet is a multi-language machine learning (ML) library to ease the
development of ML algorithms, especially for deep neural networks. Embedded in
the host language, it blends declarative symbolic expression with imperative
tensor computation. It offers auto differentiation to derive gradients. MXNet
is computation and memory efficient and runs on various heterogeneous systems,
ranging from mobile devices to distributed GPU clusters.
This paper describes both the API design and the system implementation of
MXNet, and explains how embedding of both symbolic expression and tensor
operation is handled in a unified fashion. Our preliminary experiments reveal
promising results on large scale deep neural network applications using
multiple GPU machines.
|
Predicting the top and bottom ranks of billboard songs using Machine
Learning | The music industry is a $130 billion industry. Predicting whether a song
catches the pulse of the audience impacts the industry. In this paper we
analyze language inside the lyrics of the songs using several computational
linguistic algorithms and predict whether a song would make to the top or
bottom of the billboard rankings based on the language features. We trained and
tested an SVM classifier with a radial kernel function on the linguistic
features. Results indicate that we can classify whether a song belongs to top
and bottom of the billboard charts with a precision of 0.76.
|
Predicting and visualizing psychological attributions with a deep neural
network | Judgments about personality based on facial appearance are strong effectors
in social decision making, and are known to have impact on areas from
presidential elections to jury decisions. Recent work has shown that it is
possible to predict perception of memorability, trustworthiness, intelligence
and other attributes in human face images. The most successful of these
approaches require face images expertly annotated with key facial landmarks. We
demonstrate a Convolutional Neural Network (CNN) model that is able to perform
the same task without the need for landmark features, thereby greatly
increasing efficiency. The model has high accuracy, surpassing human-level
performance in some cases. Furthermore, we use a deconvolutional approach to
visualize important features for perception of 22 attributes and demonstrate a
new method for separately visualizing positive and negative features.
|
Fixed-Point Performance Analysis of Recurrent Neural Networks | Recurrent neural networks have shown excellent performance in many
applications, however they require increased complexity in hardware or software
based implementations. The hardware complexity can be much lowered by
minimizing the word-length of weights and signals. This work analyzes the
fixed-point performance of recurrent neural networks using a retrain based
quantization method. The quantization sensitivity of each layer in RNNs is
studied, and the overall fixed-point optimization results minimizing the
capacity of weights while not sacrificing the performance are presented. A
language model and a phoneme recognition examples are used.
|
Toward a Taxonomy and Computational Models of Abnormalities in Images | The human visual system can spot an abnormal image, and reason about what
makes it strange. This task has not received enough attention in computer
vision. In this paper we study various types of atypicalities in images in a
more comprehensive way than has been done before. We propose a new dataset of
abnormal images showing a wide range of atypicalities. We design human subject
experiments to discover a coarse taxonomy of the reasons for abnormality. Our
experiments reveal three major categories of abnormality: object-centric,
scene-centric, and contextual. Based on this taxonomy, we propose a
comprehensive computational model that can predict all different types of
abnormality in images and outperform prior arts in abnormality recognition.
|
Q-Networks for Binary Vector Actions | In this paper reinforcement learning with binary vector actions was
investigated. We suggest an effective architecture of the neural networks for
approximating an action-value function with binary vector actions. The proposed
architecture approximates the action-value function by a linear function with
respect to the action vector, but is still non-linear with respect to the state
input. We show that this approximation method enables the efficient calculation
of greedy action selection and softmax action selection. Using this
architecture, we suggest an online algorithm based on Q-learning. The empirical
results in the grid world and the blocker task suggest that our approximation
architecture would be effective for the RL problems with large discrete action
sets.
|
Proposition of a Theoretical Model for Missing Data Imputation using
Deep Learning and Evolutionary Algorithms | In the last couple of decades, there has been major advancements in the
domain of missing data imputation. The techniques in the domain include amongst
others: Expectation Maximization, Neural Networks with Evolutionary Algorithms
or optimization techniques and K-Nearest Neighbor approaches to solve the
problem. The presence of missing data entries in databases render the tasks of
decision-making and data analysis nontrivial. As a result this area has
attracted a lot of research interest with the aim being to yield accurate and
time efficient and sensitive missing data imputation techniques especially when
time sensitive applications are concerned like power plants and winding
processes. In this article, considering arbitrary and monotone missing data
patterns, we hypothesize that the use of deep neural networks built using
autoencoders and denoising autoencoders in conjunction with genetic algorithms,
swarm intelligence and maximum likelihood estimator methods as novel data
imputation techniques will lead to better imputed values than existing
techniques. Also considered are the missing at random, missing completely at
random and missing not at random missing data mechanisms. We also intend to use
fuzzy logic in tandem with deep neural networks to perform the missing data
imputation tasks, as well as different building blocks for the deep neural
networks like Stacked Restricted Boltzmann Machines and Deep Belief Networks to
test our hypothesis. The motivation behind this article is the need for missing
data imputation techniques that lead to better imputed values than existing
methods with higher accuracies and lower errors.
|
Max-Pooling Dropout for Regularization of 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 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 compare max-pooling dropout and stochastic pooling,
both of which introduce stochasticity based on multinomial distributions at
pooling stage.
|
State of the Art Control of Atari Games Using Shallow Reinforcement
Learning | The recently introduced Deep Q-Networks (DQN) algorithm has gained attention
as one of the first successful combinations of deep neural networks and
reinforcement learning. Its promise was demonstrated in the Arcade Learning
Environment (ALE), a challenging framework composed of dozens of Atari 2600
games used to evaluate general competency in AI. It achieved dramatically
better results than earlier approaches, showing that its ability to learn good
representations is quite robust and general. This paper attempts to understand
the principles that underlie DQN's impressive performance and to better
contextualize its success. We systematically evaluate the importance of key
representational biases encoded by DQN's network by proposing simple linear
representations that make use of these concepts. Incorporating these
characteristics, we obtain a computationally practical feature set that
achieves competitive performance to DQN in the ALE. Besides offering insight
into the strengths and weaknesses of DQN, we provide a generic representation
for the ALE, significantly reducing the burden of learning a representation for
each game. Moreover, we also provide a simple, reproducible benchmark for the
sake of comparison to future work in the ALE.
|
Hybrid Approach for Inductive Semi Supervised Learning using Label
Propagation and Support Vector Machine | Semi supervised learning methods have gained importance in today's world
because of large expenses and time involved in labeling the unlabeled data by
human experts. The proposed hybrid approach uses SVM and Label Propagation to
label the unlabeled data. In the process, at each step SVM is trained to
minimize the error and thus improve the prediction quality. Experiments are
conducted by using SVM and logistic regression(Logreg). Results prove that SVM
performs tremendously better than Logreg. The approach is tested using 12
datasets of different sizes ranging from the order of 1000s to the order of
10000s. Results show that the proposed approach outperforms Label Propagation
by a large margin with F-measure of almost twice on average. The parallel
version of the proposed approach is also designed and implemented, the analysis
shows that the training time decreases significantly when parallel version is
used.
|
Extracting Biomolecular Interactions Using Semantic Parsing of
Biomedical Text | We advance the state of the art in biomolecular interaction extraction with
three contributions: (i) We show that deep, Abstract Meaning Representations
(AMR) significantly improve the accuracy of a biomolecular interaction
extraction system when compared to a baseline that relies solely on surface-
and syntax-based features; (ii) In contrast with previous approaches that infer
relations on a sentence-by-sentence basis, we expand our framework to enable
consistent predictions over sets of sentences (documents); (iii) We further
modify and expand a graph kernel learning framework to enable concurrent
exploitation of automatically induced AMR (semantic) and dependency structure
(syntactic) representations. Our experiments show that our approach yields
interaction extraction systems that are more robust in environments where there
is a significant mismatch between training and test conditions.
|
Creation of a Deep Convolutional Auto-Encoder in Caffe | The development of a deep (stacked) convolutional auto-encoder in the Caffe
deep learning framework is presented in this paper. We describe simple
principles which we used to create this model in Caffe. The proposed model of
convolutional auto-encoder does not have pooling/unpooling layers yet. The
results of our experimental research show comparable accuracy of dimensionality
reduction in comparison with a classic auto-encoder on the example of MNIST
dataset.
|
Risk-Constrained Reinforcement Learning with Percentile Risk Criteria | In many sequential decision-making problems one is interested in minimizing
an expected cumulative cost while taking into account \emph{risk}, i.e.,
increased awareness of events of small probability and high consequences.
Accordingly, the objective of this paper is to present efficient reinforcement
learning algorithms for risk-constrained Markov decision processes (MDPs),
where risk is represented via a chance constraint or a constraint on the
conditional value-at-risk (CVaR) of the cumulative cost. We collectively refer
to such problems as percentile risk-constrained MDPs.
Specifically, we first derive a formula for computing the gradient of the
Lagrangian function for percentile risk-constrained MDPs. Then, we devise
policy gradient and actor-critic algorithms that (1) estimate such gradient,
(2) update the policy in the descent direction, and (3) update the Lagrange
multiplier in the ascent direction. For these algorithms we prove convergence
to locally optimal policies. Finally, we demonstrate the effectiveness of our
algorithms in an optimal stopping problem and an online marketing application.
|
Approximated and User Steerable tSNE for Progressive Visual Analytics | Progressive Visual Analytics aims at improving the interactivity in existing
analytics techniques by means of visualization as well as interaction with
intermediate results. One key method for data analysis is dimensionality
reduction, for example, to produce 2D embeddings that can be visualized and
analyzed efficiently. t-Distributed Stochastic Neighbor Embedding (tSNE) is a
well-suited technique for the visualization of several high-dimensional data.
tSNE can create meaningful intermediate results but suffers from a slow
initialization that constrains its application in Progressive Visual Analytics.
We introduce a controllable tSNE approximation (A-tSNE), which trades off speed
and accuracy, to enable interactive data exploration. We offer real-time
visualization techniques, including a density-based solution and a Magic Lens
to inspect the degree of approximation. With this feedback, the user can decide
on local refinements and steer the approximation level during the analysis. We
demonstrate our technique with several datasets, in a real-world research
scenario and for the real-time analysis of high-dimensional streams to
illustrate its effectiveness for interactive data analysis.
|
Deep Attention Recurrent Q-Network | A deep learning approach to reinforcement learning led to a general learner
able to train on visual input to play a variety of arcade games at the human
and superhuman levels. Its creators at the Google DeepMind's team called the
approach: Deep Q-Network (DQN). We present an extension of DQN by "soft" and
"hard" attention mechanisms. Tests of the proposed Deep Attention Recurrent
Q-Network (DARQN) algorithm on multiple Atari 2600 games show level of
performance superior to that of DQN. Moreover, built-in attention mechanisms
allow a direct online monitoring of the training process by highlighting the
regions of the game screen the agent is focusing on when making decisions.
|
Variance Reduction for Distributed Stochastic Gradient Descent | Variance reduction (VR) methods boost the performance of stochastic gradient
descent (SGD) by enabling the use of larger, constant stepsizes and preserving
linear convergence rates. However, current variance reduced SGD methods require
either high memory usage or an exact gradient computation (using the entire
dataset) at the end of each epoch. This limits the use of VR methods in
practical distributed settings. In this paper, we propose a variance reduction
method, called VR-lite, that does not require full gradient computations or
extra storage. We explore distributed synchronous and asynchronous variants
that are scalable and remain stable with low communication frequency. We
empirically compare both the sequential and distributed algorithms to
state-of-the-art stochastic optimization methods, and find that our proposed
algorithms perform favorably to other stochastic methods.
|
Generating News Headlines with Recurrent Neural Networks | We describe an application of an encoder-decoder recurrent neural network
with LSTM units and attention to generating headlines from the text of news
articles. We find that the model is quite effective at concisely paraphrasing
news articles. Furthermore, we study how the neural network decides which input
words to pay attention to, and specifically we identify the function of the
different neurons in a simplified attention mechanism. Interestingly, our
simplified attention mechanism performs better that the more complex attention
mechanism on a held out set of articles.
|
Similarity Learning via Adaptive Regression and Its Application to Image
Retrieval | We study the problem of similarity learning and its application to image
retrieval with large-scale data. The similarity between pairs of images can be
measured by the distances between their high dimensional representations, and
the problem of learning the appropriate similarity is often addressed by
distance metric learning. However, distance metric learning requires the
learned metric to be a PSD matrix, which is computational expensive and not
necessary for retrieval ranking problem. On the other hand, the bilinear model
is shown to be more flexible for large-scale image retrieval task, hence, we
adopt it to learn a matrix for estimating pairwise similarities under the
regression framework. By adaptively updating the target matrix in regression,
we can mimic the hinge loss, which is more appropriate for similarity learning
problem. Although the regression problem can have the closed-form solution, the
computational cost can be very expensive. The computational challenges come
from two aspects: the number of images can be very large and image features
have high dimensionality. We address the first challenge by compressing the
data by a randomized algorithm with the theoretical guarantee. For the high
dimensional issue, we address it by taking low rank assumption and applying
alternating method to obtain the partial matrix, which has a global optimal
solution. Empirical studies on real world image datasets (i.e., Caltech and
ImageNet) demonstrate the effectiveness and efficiency of the proposed method.
|
Large Scale Distributed Semi-Supervised Learning Using Streaming
Approximation | Traditional graph-based semi-supervised learning (SSL) approaches, even
though widely applied, are not suited for massive data and large label
scenarios since they scale linearly with the number of edges $|E|$ and distinct
labels $m$. To deal with the large label size problem, recent works propose
sketch-based methods to approximate the distribution on labels per node thereby
achieving a space reduction from $O(m)$ to $O(\log m)$, under certain
conditions. In this paper, we present a novel streaming graph-based SSL
approximation that captures the sparsity of the label distribution and ensures
the algorithm propagates labels accurately, and further reduces the space
complexity per node to $O(1)$. We also provide a distributed version of the
algorithm that scales well to large data sizes. Experiments on real-world
datasets demonstrate that the new method achieves better performance than
existing state-of-the-art algorithms with significant reduction in memory
footprint. We also study different graph construction mechanisms for natural
language applications and propose a robust graph augmentation strategy trained
using state-of-the-art unsupervised deep learning architectures that yields
further significant quality gains.
|
Explaining reviews and ratings with PACO: Poisson Additive Co-Clustering | Understanding a user's motivations provides valuable information beyond the
ability to recommend items. Quite often this can be accomplished by perusing
both ratings and review texts, since it is the latter where the reasoning for
specific preferences is explicitly expressed.
Unfortunately matrix factorization approaches to recommendation result in
large, complex models that are difficult to interpret and give recommendations
that are hard to clearly explain to users. In contrast, in this paper, we
attack this problem through succinct additive co-clustering. We devise a novel
Bayesian technique for summing co-clusterings of Poisson distributions. With
this novel technique we propose a new Bayesian model for joint collaborative
filtering of ratings and text reviews through a sum of simple co-clusterings.
The simple structure of our model yields easily interpretable recommendations.
Even with a simple, succinct structure, our model outperforms competitors in
terms of predicting ratings with reviews.
|
Rademacher Complexity of the Restricted Boltzmann Machine | Boltzmann machine, as a fundamental construction block of deep belief network
and deep Boltzmann machines, is widely used in deep learning community and
great success has been achieved. However, theoretical understanding of many
aspects of it is still far from clear. In this paper, we studied the Rademacher
complexity of both the asymptotic restricted Boltzmann machine and the
practical implementation with single-step contrastive divergence (CD-1)
procedure. Our results disclose the fact that practical implementation training
procedure indeed increased the Rademacher complexity of restricted Boltzmann
machines. A further research direction might be the investigation of the VC
dimension of a compositional function used in the CD-1 procedure.
|
Thinking Required | There exists a theory of a single general-purpose learning algorithm which
could explain the principles its operation. It assumes the initial rough
architecture, a small library of simple innate circuits which are prewired at
birth. and proposes that all significant mental algorithms are learned. Given
current understanding and observations, this paper reviews and lists the
ingredients of such an algorithm from architectural and functional
perspectives.
|
Fast Optimization Algorithm on Riemannian Manifolds and Its Application
in Low-Rank Representation | The paper addresses the problem of optimizing a class of composite functions
on Riemannian manifolds and a new first order optimization algorithm (FOA) with
a fast convergence rate is proposed. Through the theoretical analysis for FOA,
it has been proved that the algorithm has quadratic convergence. The
experiments in the matrix completion task show that FOA has better performance
than other first order optimization methods on Riemannian manifolds. A fast
subspace pursuit method based on FOA is proposed to solve the low-rank
representation model based on augmented Lagrange method on the low rank matrix
variety. Experimental results on synthetic and real data sets are presented to
demonstrate that both FOA and SP-RPRG(ALM) can achieve superior performance in
terms of faster convergence and higher accuracy.
|
A Novel Approach to Distributed Multi-Class SVM | With data sizes constantly expanding, and with classical machine learning
algorithms that analyze such data requiring larger and larger amounts of
computation time and storage space, the need to distribute computation and
memory requirements among several computers has become apparent. Although
substantial work has been done in developing distributed binary SVM algorithms
and multi-class SVM algorithms individually, the field of multi-class
distributed SVMs remains largely unexplored. This research proposes a novel
algorithm that implements the Support Vector Machine over a multi-class dataset
and is efficient in a distributed environment (here, Hadoop). The idea is to
divide the dataset into half recursively and thus compute the optimal Support
Vector Machine for this half during the training phase, much like a divide and
conquer approach. While testing, this structure has been effectively exploited
to significantly reduce the prediction time. Our algorithm has shown better
computation time during the prediction phase than the traditional sequential
SVM methods (One vs. One, One vs. Rest) and out-performs them as the size of
the dataset grows. This approach also classifies the data with higher accuracy
than the traditional multi-class algorithms.
|
Jointly Modeling Topics and Intents with Global Order Structure | Modeling document structure is of great importance for discourse analysis and
related applications. The goal of this research is to capture the document
intent structure by modeling documents as a mixture of topic words and
rhetorical words. While the topics are relatively unchanged through one
document, the rhetorical functions of sentences usually change following
certain orders in discourse. We propose GMM-LDA, a topic modeling based
Bayesian unsupervised model, to analyze the document intent structure
cooperated with order information. Our model is flexible that has the ability
to combine the annotations and do supervised learning. Additionally, entropic
regularization can be introduced to model the significant divergence between
topics and intents. We perform experiments in both unsupervised and supervised
settings, results show the superiority of our model over several
state-of-the-art baselines.
|
How to Discount Deep Reinforcement Learning: Towards New Dynamic
Strategies | Using deep neural nets as function approximator for reinforcement learning
tasks have recently been shown to be very powerful for solving problems
approaching real-world complexity. Using these results as a benchmark, we
discuss the role that the discount factor may play in the quality of the
learning process of a deep Q-network (DQN). When the discount factor
progressively increases up to its final value, we empirically show that it is
possible to significantly reduce the number of learning steps. When used in
conjunction with a varying learning rate, we empirically show that it
outperforms original DQN on several experiments. We relate this phenomenon with
the instabilities of neural networks when they are used in an approximate
Dynamic Programming setting. We also describe the possibility to fall within a
local optimum during the learning process, thus connecting our discussion with
the exploration/exploitation dilemma.
|
Discriminative Nonparametric Latent Feature Relational Models with Data
Augmentation | We present a discriminative nonparametric latent feature relational model
(LFRM) for link prediction to automatically infer the dimensionality of latent
features. Under the generic RegBayes (regularized Bayesian inference)
framework, we handily incorporate the prediction loss with probabilistic
inference of a Bayesian model; set distinct regularization parameters for
different types of links to handle the imbalance issue in real networks; and
unify the analysis of both the smooth logistic log-loss and the piecewise
linear hinge loss. For the nonconjugate posterior inference, we present a
simple Gibbs sampler via data augmentation, without making restricting
assumptions as done in variational methods. We further develop an approximate
sampler using stochastic gradient Langevin dynamics to handle large networks
with hundreds of thousands of entities and millions of links, orders of
magnitude larger than what existing LFRM models can process. Extensive studies
on various real networks show promising performance.
|
Risk Minimization in Structured Prediction using Orbit Loss | We introduce a new surrogate loss function called orbit loss in the
structured prediction framework, which has good theoretical and practical
advantages. While the orbit loss is not convex, it has a simple analytical
gradient and a simple perceptron-like learning rule. We analyze the new loss
theoretically and state a PAC-Bayesian generalization bound. We also prove that
the new loss is consistent in the strong sense; namely, the risk achieved by
the set of the trained parameters approaches the infimum risk achievable by any
linear decoder over the given features. Methods that are aimed at risk
minimization, such as the structured ramp loss, the structured probit loss and
the direct loss minimization require at least two inference operations per
training iteration. In this sense, the orbit loss is more efficient as it
requires only one inference operation per training iteration, while yields
similar performance. We conclude the paper with an empirical comparison of the
proposed loss function to the structured hinge loss, the structured ramp loss,
the structured probit loss and the direct loss minimization method on several
benchmark datasets and tasks.
|
Clustering by Deep Nearest Neighbor Descent (D-NND): A Density-based
Parameter-Insensitive Clustering Method | Most density-based clustering methods largely rely on how well the underlying
density is estimated. However, density estimation itself is also a challenging
problem, especially the determination of the kernel bandwidth. A large
bandwidth could lead to the over-smoothed density estimation in which the
number of density peaks could be less than the true clusters, while a small
bandwidth could lead to the under-smoothed density estimation in which spurious
density peaks, or called the "ripple noise", would be generated in the
estimated density. In this paper, we propose a density-based hierarchical
clustering method, called the Deep Nearest Neighbor Descent (D-NND), which
could learn the underlying density structure layer by layer and capture the
cluster structure at the same time. The over-smoothed density estimation could
be largely avoided and the negative effect of the under-estimated cases could
be also largely reduced. Overall, D-NND presents not only the strong capability
of discovering the underlying cluster structure but also the remarkable
reliability due to its insensitivity to parameters.
|
Obtaining A Linear Combination of the Principal Components of a Matrix
on Quantum Computers | Principal component analysis is a multivariate statistical method frequently
used in science and engineering to reduce the dimension of a problem or extract
the most significant features from a dataset. In this paper, using a similar
notion to the quantum counting, we show how to apply the amplitude
amplification together with the phase estimation algorithm to an operator in
order to procure the eigenvectors of the operator associated to the eigenvalues
defined in the range $\left[a, b\right]$, where $a$ and $b$ are real and $0
\leq a \leq b \leq 1$. This makes possible to obtain a combination of the
eigenvectors associated to the largest eigenvalues and so can be used to do
principal component analysis on quantum computers.
|
A Large Dataset to Train Convolutional Networks for Disparity, Optical
Flow, and Scene Flow Estimation | Recent work has shown that optical flow estimation can be formulated as a
supervised learning task and can be successfully solved with convolutional
networks. Training of the so-called FlowNet was enabled by a large
synthetically generated dataset. The present paper extends the concept of
optical flow estimation via convolutional networks to disparity and scene flow
estimation. To this end, we propose three synthetic stereo video datasets with
sufficient realism, variation, and size to successfully train large networks.
Our datasets are the first large-scale datasets to enable training and
evaluating scene flow methods. Besides the datasets, we present a convolutional
network for real-time disparity estimation that provides state-of-the-art
results. By combining a flow and disparity estimation network and training it
jointly, we demonstrate the first scene flow estimation with a convolutional
network.
|
The Teaching Dimension of Linear Learners | Teaching dimension is a learning theoretic quantity that specifies the
minimum training set size to teach a target model to a learner. Previous
studies on teaching dimension focused on version-space learners which maintain
all hypotheses consistent with the training data, and cannot be applied to
modern machine learners which select a specific hypothesis via optimization.
This paper presents the first known teaching dimension for ridge regression,
support vector machines, and logistic regression. We also exhibit optimal
training sets that match these teaching dimensions. Our approach generalizes to
other linear learners.
|
Pseudo-Bayesian Robust PCA: Algorithms and Analyses | Commonly used in computer vision and other applications, robust PCA
represents an algorithmic attempt to reduce the sensitivity of classical PCA to
outliers. The basic idea is to learn a decomposition of some data matrix of
interest into low rank and sparse components, the latter representing unwanted
outliers. Although the resulting optimization problem is typically NP-hard,
convex relaxations provide a computationally-expedient alternative with
theoretical support. However, in practical regimes performance guarantees break
down and a variety of non-convex alternatives, including Bayesian-inspired
models, have been proposed to boost estimation quality. Unfortunately though,
without additional a priori knowledge none of these methods can significantly
expand the critical operational range such that exact principal subspace
recovery is possible. Into this mix we propose a novel pseudo-Bayesian
algorithm that explicitly compensates for design weaknesses in many existing
non-convex approaches leading to state-of-the-art performance with a sound
analytical foundation. Surprisingly, our algorithm can even outperform convex
matrix completion despite the fact that the latter is provided with perfect
knowledge of which entries are not corrupted.
|
Fast spectral algorithms from sum-of-squares proofs: tensor
decomposition and planted sparse vectors | We consider two problems that arise in machine learning applications: the
problem of recovering a planted sparse vector in a random linear subspace and
the problem of decomposing a random low-rank overcomplete 3-tensor. For both
problems, the best known guarantees are based on the sum-of-squares method. We
develop new algorithms inspired by analyses of the sum-of-squares method. Our
algorithms achieve the same or similar guarantees as sum-of-squares for these
problems but the running time is significantly faster.
For the planted sparse vector problem, we give an algorithm with running time
nearly linear in the input size that approximately recovers a planted sparse
vector with up to constant relative sparsity in a random subspace of $\mathbb
R^n$ of dimension up to $\tilde \Omega(\sqrt n)$. These recovery guarantees
match the best known ones of Barak, Kelner, and Steurer (STOC 2014) up to
logarithmic factors.
For tensor decomposition, we give an algorithm with running time close to
linear in the input size (with exponent $\approx 1.086$) that approximately
recovers a component of a random 3-tensor over $\mathbb R^n$ of rank up to
$\tilde \Omega(n^{4/3})$. The best previous algorithm for this problem due to
Ge and Ma (RANDOM 2015) works up to rank $\tilde \Omega(n^{3/2})$ but requires
quasipolynomial time.
|
Online Crowdsourcing | With the success of modern internet based platform, such as Amazon Mechanical
Turk, it is now normal to collect a large number of hand labeled samples from
non-experts. The Dawid- Skene algorithm, which is based on Expectation-
Maximization update, has been widely used for inferring the true labels from
noisy crowdsourced labels. However, Dawid-Skene scheme requires all the data to
perform each EM iteration, and can be infeasible for streaming data or large
scale data. In this paper, we provide an online version of Dawid- Skene
algorithm that only requires one data frame for each iteration. Further, we
prove that under mild conditions, the online Dawid-Skene scheme with projection
converges to a stationary point of the marginal log-likelihood of the observed
data. Our experiments demonstrate that the online Dawid- Skene scheme achieves
state of the art performance comparing with other methods based on the Dawid-
Skene scheme.
|
Online Gradient Descent in Function Space | In many problems in machine learning and operations research, we need to
optimize a function whose input is a random variable or a probability density
function, i.e. to solve optimization problems in an infinite dimensional space.
On the other hand, online learning has the advantage of dealing with streaming
examples, and better model a changing environ- ment. In this paper, we extend
the celebrated online gradient descent algorithm to Hilbert spaces (function
spaces), and analyze the convergence guarantee of the algorithm. Finally, we
demonstrate that our algorithms can be useful in several important problems.
|
Learning Discrete Bayesian Networks from Continuous Data | Learning Bayesian networks from raw data can help provide insights into the
relationships between variables. While real data often contains a mixture of
discrete and continuous-valued variables, many Bayesian network structure
learning algorithms assume all random variables are discrete. Thus, continuous
variables are often discretized when learning a Bayesian network. However, the
choice of discretization policy has significant impact on the accuracy, speed,
and interpretability of the resulting models. This paper introduces a
principled Bayesian discretization method for continuous variables in Bayesian
networks with quadratic complexity instead of the cubic complexity of other
standard techniques. Empirical demonstrations show that the proposed method is
superior to the established minimum description length algorithm. In addition,
this paper shows how to incorporate existing methods into the structure
learning process to discretize all continuous variables and simultaneously
learn Bayesian network structures.
|
Explaining NonLinear Classification Decisions with Deep Taylor
Decomposition | Nonlinear methods such as Deep Neural Networks (DNNs) are the gold standard
for various challenging machine learning problems, e.g., image classification,
natural language processing or human action recognition. Although these methods
perform impressively well, they have a significant disadvantage, the lack of
transparency, limiting the interpretability of the solution and thus the scope
of application in practice. Especially DNNs act as black boxes due to their
multilayer nonlinear structure. In this paper we introduce a novel methodology
for interpreting generic multilayer neural networks by decomposing the network
classification decision into contributions of its input elements. Although our
focus is on image classification, the method is applicable to a broad set of
input data, learning tasks and network architectures. Our method is based on
deep Taylor decomposition and efficiently utilizes the structure of the network
by backpropagating the explanations from the output to the input layer. We
evaluate the proposed method empirically on the MNIST and ILSVRC data sets.
|
Deep Exemplar 2D-3D Detection by Adapting from Real to Rendered Views | This paper presents an end-to-end convolutional neural network (CNN) for
2D-3D exemplar detection. We demonstrate that the ability to adapt the features
of natural images to better align with those of CAD rendered views is critical
to the success of our technique. We show that the adaptation can be learned by
compositing rendered views of textured object models on natural images. Our
approach can be naturally incorporated into a CNN detection pipeline and
extends the accuracy and speed benefits from recent advances in deep learning
to 2D-3D exemplar detection. We applied our method to two tasks: instance
detection, where we evaluated on the IKEA dataset, and object category
detection, where we out-perform Aubry et al. for "chair" detection on a subset
of the Pascal VOC dataset.
|
Deep Learning for Single and Multi-Session i-Vector Speaker Recognition | The promising performance of Deep Learning (DL) in speech recognition has
motivated the use of DL in other speech technology applications such as speaker
recognition. Given i-vectors as inputs, the authors proposed an impostor
selection algorithm and a universal model adaptation process in a hybrid system
based on Deep Belief Networks (DBN) and Deep Neural Networks (DNN) to
discriminatively model each target speaker. In order to have more insight into
the behavior of DL techniques in both single and multi-session speaker
enrollment tasks, some experiments have been carried out in this paper in both
scenarios. Additionally, the parameters of the global model, referred to as
universal DBN (UDBN), are normalized before adaptation. UDBN normalization
facilitates training DNNs specifically with more than one hidden layer.
Experiments are performed on the NIST SRE 2006 corpus. It is shown that the
proposed impostor selection algorithm and UDBN adaptation process enhance the
performance of conventional DNNs 8-20 % and 16-20 % in terms of EER for the
single and multi-session tasks, respectively. In both scenarios, the proposed
architectures outperform the baseline systems obtaining up to 17 % reduction in
EER.
|
Speeding Up Distributed Machine Learning Using Codes | Codes are widely used in many engineering applications to offer robustness
against noise. In large-scale systems there are several types of noise that can
affect the performance of distributed machine learning algorithms -- straggler
nodes, system failures, or communication bottlenecks -- but there has been
little interaction cutting across codes, machine learning, and distributed
systems. In this work, we provide theoretical insights on how coded solutions
can achieve significant gains compared to uncoded ones. We focus on two of the
most basic building blocks of distributed learning algorithms: matrix
multiplication and data shuffling. For matrix multiplication, we use codes to
alleviate the effect of stragglers, and show that if the number of homogeneous
workers is $n$, and the runtime of each subtask has an exponential tail, coded
computation can speed up distributed matrix multiplication by a factor of $\log
n$. For data shuffling, we use codes to reduce communication bottlenecks,
exploiting the excess in storage. We show that when a constant fraction
$\alpha$ of the data matrix can be cached at each worker, and $n$ is the number
of workers, \emph{coded shuffling} reduces the communication cost by a factor
of $(\alpha + \frac{1}{n})\gamma(n)$ compared to uncoded shuffling, where
$\gamma(n)$ is the ratio of the cost of unicasting $n$ messages to $n$ users to
multicasting a common message (of the same size) to $n$ users. For instance,
$\gamma(n) \simeq n$ if multicasting a message to $n$ users is as cheap as
unicasting a message to one user. We also provide experiment results,
corroborating our theoretical gains of the coded algorithms.
|
Reinforcement Control with Hierarchical Backpropagated Adaptive Critics | Present incremental learning methods are limited in the ability to achieve
reliable credit assignment over a large number time steps (or events). However,
this situation is typical for cases where the dynamical system to be controlled
requires relatively frequent control updates in order to maintain stability or
robustness yet has some action-consequences which must be established over
relatively long periods of time. To address this problem, the learning
capabilities of a control architecture comprised of two Backpropagated Adaptive
Critics (BACs) in a two-level hierarchy with continuous actions are explored.
The high-level BAC updates less frequently than the low-level BAC and controls
the latter to some degree. The response of the low-level to high-level signals
can either be determined a priori or it can emerge during learning. A general
approach called Response Induction Learning is introduced to address the latter
case.
|
Distributed Training of Deep Neural Networks with Theoretical Analysis:
Under SSP Setting | We propose a distributed approach to train deep neural networks (DNNs), which
has guaranteed convergence theoretically and great scalability empirically:
close to 6 times faster on instance of ImageNet data set when run with 6
machines. The proposed scheme is close to optimally scalable in terms of number
of machines, and guaranteed to converge to the same optima as the undistributed
setting. The convergence and scalability of the distributed setting is shown
empirically across different datasets (TIMIT and ImageNet) and machine learning
tasks (image classification and phoneme extraction). The convergence analysis
provides novel insights into this complex learning scheme, including: 1)
layerwise convergence, and 2) convergence of the weights in probability.
|
Window-Object Relationship Guided Representation Learning for Generic
Object Detections | In existing works that learn representation for object detection, the
relationship between a candidate window and the ground truth bounding box of an
object is simplified by thresholding their overlap. This paper shows
information loss in this simplification and picks up the relative location/size
information discarded by thresholding. We propose a representation learning
pipeline to use the relationship as supervision for improving the learned
representation in object detection. Such relationship is not limited to object
of the target category, but also includes surrounding objects of other
categories. We show that image regions with multiple contexts and multiple
rotations are effective in capturing such relationship during the
representation learning process and in handling the semantic and visual
variation caused by different window-object configurations. Experimental
results show that the representation learned by our approach can improve the
object detection accuracy by 6.4% in mean average precision (mAP) on
ILSVRC2014. On the challenging ILSVRC2014 test dataset, 48.6% mAP is achieved
by our single model and it is the best among published results. On PASCAL VOC,
it outperforms the state-of-the-art result of Fast RCNN by 3.3% in absolute
mAP.
|
Perfect Recovery Conditions For Non-Negative Sparse Modeling | Sparse modeling has been widely and successfully used in many applications
such as computer vision, machine learning, and pattern recognition. Accompanied
with those applications, significant research has studied the theoretical
limits and algorithm design for convex relaxations in sparse modeling. However,
theoretical analyses on non-negative versions of sparse modeling are limited in
the literature either to a noiseless setting or a scenario with a specific
statistical noise model such as Gaussian noise. This paper studies the
performance of non-negative sparse modeling in a more general scenario where
the observed signals have an unknown arbitrary distortion, especially focusing
on non-negativity constrained and L1-penalized least squares, and gives an
exact bound for which this problem can recover the correct signal elements. We
pose two conditions to guarantee the correct signal recovery: minimum
coefficient condition (MCC) and nonlinearity vs. subset coherence condition
(NSCC). The former defines the minimum weight for each of the correct atoms
present in the signal and the latter defines the tolerable deviation from the
linear model relative to the positive subset coherence (PSC), a novel type of
"coherence" metric. We provide rigorous performance guarantees based on these
conditions and experimentally verify their precise predictive power in a
hyperspectral data unmixing application.
|
A Novel Regularized Principal Graph Learning Framework on Explicit Graph
Representation | Many scientific datasets are of high dimension, and the analysis usually
requires visual manipulation by retaining the most important structures of
data. Principal curve is a widely used approach for this purpose. However, many
existing methods work only for data with structures that are not
self-intersected, which is quite restrictive for real applications. A few
methods can overcome the above problem, but they either require complicated
human-made rules for a specific task with lack of convergence guarantee and
adaption flexibility to different tasks, or cannot obtain explicit structures
of data. To address these issues, we develop a new regularized principal graph
learning framework that captures the local information of the underlying graph
structure based on reversed graph embedding. As showcases, models that can
learn a spanning tree or a weighted undirected $\ell_1$ graph are proposed, and
a new learning algorithm is developed that learns a set of principal points and
a graph structure from data, simultaneously. The new algorithm is simple with
guaranteed convergence. We then extend the proposed framework to deal with
large-scale data. Experimental results on various synthetic and six real world
datasets show that the proposed method compares favorably with baselines and
can uncover the underlying structure correctly.
|
Affinity CNN: Learning Pixel-Centric Pairwise Relations for
Figure/Ground Embedding | Spectral embedding provides a framework for solving perceptual organization
problems, including image segmentation and figure/ground organization. From an
affinity matrix describing pairwise relationships between pixels, it clusters
pixels into regions, and, using a complex-valued extension, orders pixels
according to layer. We train a convolutional neural network (CNN) to directly
predict the pairwise relationships that define this affinity matrix. Spectral
embedding then resolves these predictions into a globally-consistent
segmentation and figure/ground organization of the scene. Experiments
demonstrate significant benefit to this direct coupling compared to prior works
which use explicit intermediate stages, such as edge detection, on the pathway
from image to affinities. Our results suggest spectral embedding as a powerful
alternative to the conditional random field (CRF)-based globalization schemes
typically coupled to deep neural networks.
|
Bigger Buffer k-d Trees on Multi-Many-Core Systems | A buffer k-d tree is a k-d tree variant for massively-parallel nearest
neighbor search. While providing valuable speed-ups on modern many-core devices
in case both a large number of reference and query points are given, buffer k-d
trees are limited by the amount of points that can fit on a single device. In
this work, we show how to modify the original data structure and the associated
workflow to make the overall approach capable of dealing with massive data
sets. We further provide a simple yet efficient way of using multiple devices
given in a single workstation. The applicability of the modified framework is
demonstrated in the context of astronomy, a field that is faced with huge
amounts of data.
|
Multi-Player Bandits -- a Musical Chairs Approach | We consider a variant of the stochastic multi-armed bandit problem, where
multiple players simultaneously choose from the same set of arms and may
collide, receiving no reward. This setting has been motivated by problems
arising in cognitive radio networks, and is especially challenging under the
realistic assumption that communication between players is limited. We provide
a communication-free algorithm (Musical Chairs) which attains constant regret
with high probability, as well as a sublinear-regret, communication-free
algorithm (Dynamic Musical Chairs) for the more difficult setting of players
dynamically entering and leaving throughout the game. Moreover, both algorithms
do not require prior knowledge of the number of players. To the best of our
knowledge, these are the first communication-free algorithms with these types
of formal guarantees. We also rigorously compare our algorithms to previous
works, and complement our theoretical findings with experiments.
|
Where You Are Is Who You Are: User Identification by Matching Statistics | Most users of online services have unique behavioral or usage patterns. These
behavioral patterns can be exploited to identify and track users by using only
the observed patterns in the behavior. We study the task of identifying users
from statistics of their behavioral patterns. Specifically, we focus on the
setting in which we are given histograms of users' data collected during two
different experiments. We assume that, in the first dataset, the users'
identities are anonymized or hidden and that, in the second dataset, their
identities are known. We study the task of identifying the users by matching
the histograms of their data in the first dataset with the histograms from the
second dataset. In recent works, the optimal algorithm for this user
identification task is introduced. In this paper, we evaluate the effectiveness
of this method on three different types of datasets and in multiple scenarios.
Using datasets such as call data records, web browsing histories, and GPS
trajectories, we show that a large fraction of users can be easily identified
given only histograms of their data; hence these histograms can act as users'
fingerprints. We also verify that simultaneous identification of users achieves
better performance compared to one-by-one user identification. We show that
using the optimal method for identification gives higher identification
accuracy than heuristics-based approaches in practical scenarios. The accuracy
obtained under this optimal method can thus be used to quantify the maximum
level of user identification that is possible in such settings. We show that
the key factors affecting the accuracy of the optimal identification algorithm
are the duration of the data collection, the number of users in the anonymized
dataset, and the resolution of the dataset. We analyze the effectiveness of
k-anonymization in resisting user identification attacks on these datasets.
|
Efficient Distributed SGD with Variance Reduction | Stochastic Gradient Descent (SGD) has become one of the most popular
optimization methods for training machine learning models on massive datasets.
However, SGD suffers from two main drawbacks: (i) The noisy gradient updates
have high variance, which slows down convergence as the iterates approach the
optimum, and (ii) SGD scales poorly in distributed settings, typically
experiencing rapidly decreasing marginal benefits as the number of workers
increases. In this paper, we propose a highly parallel method, CentralVR, that
uses error corrections to reduce the variance of SGD gradient updates, and
scales linearly with the number of worker nodes. CentralVR enjoys low iteration
complexity, provably linear convergence rates, and exhibits linear performance
gains up to hundreds of cores for massive datasets. We compare CentralVR to
state-of-the-art parallel stochastic optimization methods on a variety of
models and datasets, and find that our proposed methods exhibit stronger
scaling than other SGD variants.
|
Partial Reinitialisation for Optimisers | Heuristic optimisers which search for an optimal configuration of variables
relative to an objective function often get stuck in local optima where the
algorithm is unable to find further improvement. The standard approach to
circumvent this problem involves periodically restarting the algorithm from
random initial configurations when no further improvement can be found. We
propose a method of partial reinitialization, whereby, in an attempt to find a
better solution, only sub-sets of variables are re-initialised rather than the
whole configuration. Much of the information gained from previous runs is hence
retained. This leads to significant improvements in the quality of the solution
found in a given time for a variety of optimisation problems in machine
learning.
|
Gated networks: an inventory | Gated networks are networks that contain gating connections, in which the
outputs of at least two neurons are multiplied. Initially, gated networks were
used to learn relationships between two input sources, such as pixels from two
images. More recently, they have been applied to learning activity recognition
or multi-modal representations. The aims of this paper are threefold: 1) to
explain the basic computations in gated networks to the non-expert, while
adopting a standpoint that insists on their symmetric nature. 2) to serve as a
quick reference guide to the recent literature, by providing an inventory of
applications of these networks, as well as recent extensions to the basic
architecture. 3) to suggest future research directions and applications.
|
Norm-Free Radon-Nikodym Approach to Machine Learning | For Machine Learning (ML) classification problem, where a vector of
$\mathbf{x}$--observations (values of attributes) is mapped to a single $y$
value (class label), a generalized Radon--Nikodym type of solution is proposed.
Quantum--mechanics --like probability states $\psi^2(\mathbf{x})$ are
considered and "Cluster Centers", corresponding to the extremums of
$<y\psi^2(\mathbf{x})>/<\psi^2(\mathbf{x})>$, are found from generalized
eigenvalues problem. The eigenvalues give possible $y^{[i]}$ outcomes and
corresponding to them eigenvectors $\psi^{[i]}(\mathbf{x})$ define "Cluster
Centers". The projection of a $\psi$ state, localized at given $\mathbf{x}$ to
classify, on these eigenvectors define the probability of $y^{[i]}$ outcome,
thus avoiding using a norm ($L^2$ or other types), required for "quality
criteria" in a typical Machine Learning technique. A coverage of each `Cluster
Center" is calculated, what potentially allows to separate system properties
(described by $y^{[i]}$ outcomes) and system testing conditions (described by
$C^{[i]}$ coverage). As an example of such application $y$ distribution
estimator is proposed in a form of pairs $(y^{[i]},C^{[i]})$, that can be
considered as Gauss quadratures generalization. This estimator allows to
perform $y$ probability distribution estimation in a strongly non--Gaussian
case.
|
Convolutional Monte Carlo Rollouts in Go | In this work, we present a MCTS-based Go-playing program which uses
convolutional networks in all parts. Our method performs MCTS in batches,
explores the Monte Carlo search tree using Thompson sampling and a
convolutional network, and evaluates convnet-based rollouts on the GPU. We
achieve strong win rates against open source Go programs and attain competitive
results against state of the art convolutional net-based Go-playing programs.
|
Boosted Sparse Non-linear Distance Metric Learning | This paper proposes a boosting-based solution addressing metric learning
problems for high-dimensional data. Distance measures have been used as natural
measures of (dis)similarity and served as the foundation of various learning
methods. The efficiency of distance-based learning methods heavily depends on
the chosen distance metric. With increasing dimensionality and complexity of
data, however, traditional metric learning methods suffer from poor scalability
and the limitation due to linearity as the true signals are usually embedded
within a low-dimensional nonlinear subspace. In this paper, we propose a
nonlinear sparse metric learning algorithm via boosting. We restructure a
global optimization problem into a forward stage-wise learning of weak learners
based on a rank-one decomposition of the weight matrix in the Mahalanobis
distance metric. A gradient boosting algorithm is devised to obtain a sparse
rank-one update of the weight matrix at each step. Nonlinear features are
learned by a hierarchical expansion of interactions incorporated within the
boosting algorithm. Meanwhile, an early stopping rule is imposed to control the
overall complexity of the learned metric. As a result, our approach guarantees
three desirable properties of the final metric: positive semi-definiteness, low
rank and element-wise sparsity. Numerical experiments show that our learning
model compares favorably with the state-of-the-art methods in the current
literature of metric learning.
|
Predicting proximity with ambient mobile sensors for non-invasive health
diagnostics | Modern smart phones are becoming helpful in the areas of Internet-Of-Things
(IoT) and ambient health intelligence. By learning data from several mobile
sensors, we detect nearness of the human body to a mobile device in a
three-dimensional space with no physical contact with the device for
non-invasive health diagnostics. We show that the human body generates wave
patterns that interact with other naturally occurring ambient signals that
could be measured by mobile sensors, such as, temperature, humidity, magnetic
field, acceleration, gravity, and light. This interaction consequentially
alters the patterns of the naturally occurring signals, and thus, exhibits
characteristics that could be learned to predict the nearness of the human body
to a mobile device, hence provide diagnostic information for medical
practitioners. Our prediction technique achieved 88.75% accuracy and 88.3%
specificity.
|
A Unified Approach to Error Bounds for Structured Convex Optimization
Problems | Error bounds, which refer to inequalities that bound the distance of vectors
in a test set to a given set by a residual function, have proven to be
extremely useful in analyzing the convergence rates of a host of iterative
methods for solving optimization problems. In this paper, we present a new
framework for establishing error bounds for a class of structured convex
optimization problems, in which the objective function is the sum of a smooth
convex function and a general closed proper convex function. Such a class
encapsulates not only fairly general constrained minimization problems but also
various regularized loss minimization formulations in machine learning, signal
processing, and statistics. Using our framework, we show that a number of
existing error bound results can be recovered in a unified and transparent
manner. To further demonstrate the power of our framework, we apply it to a
class of nuclear-norm regularized loss minimization problems and establish a
new error bound for this class under a strict complementarity-type regularity
condition. We then complement this result by constructing an example to show
that the said error bound could fail to hold without the regularity condition.
Consequently, we obtain a rather complete answer to a question raised by Tseng.
We believe that our approach will find further applications in the study of
error bounds for structured convex optimization problems.
|
Distilling Knowledge from Deep Networks with Applications to Healthcare
Domain | Exponential growth in Electronic Healthcare Records (EHR) has resulted in new
opportunities and urgent needs for discovery of meaningful data-driven
representations and patterns of diseases in Computational Phenotyping research.
Deep Learning models have shown superior performance for robust prediction in
computational phenotyping tasks, but suffer from the issue of model
interpretability which is crucial for clinicians involved in decision-making.
In this paper, we introduce a novel knowledge-distillation approach called
Interpretable Mimic Learning, to learn interpretable phenotype features for
making robust prediction while mimicking the performance of deep learning
models. Our framework uses Gradient Boosting Trees to learn interpretable
features from deep learning models such as Stacked Denoising Autoencoder and
Long Short-Term Memory. Exhaustive experiments on a real-world clinical
time-series dataset show that our method obtains similar or better performance
than the deep learning models, and it provides interpretable phenotypes for
clinical decision making.
|
Words are not Equal: Graded Weighting Model for building Composite
Document Vectors | Despite the success of distributional semantics, composing phrases from word
vectors remains an important challenge. Several methods have been tried for
benchmark tasks such as sentiment classification, including word vector
averaging, matrix-vector approaches based on parsing, and on-the-fly learning
of paragraph vectors. Most models usually omit stop words from the composition.
Instead of such an yes-no decision, we consider several graded schemes where
words are weighted according to their discriminatory relevance with respect to
its use in the document (e.g., idf). Some of these methods (particularly
tf-idf) are seen to result in a significant improvement in performance over
prior state of the art. Further, combining such approaches into an ensemble
based on alternate classifiers such as the RNN model, results in an 1.6%
performance improvement on the standard IMDB movie review dataset, and a 7.01%
improvement on Amazon product reviews. Since these are language free models and
can be obtained in an unsupervised manner, they are of interest also for
under-resourced languages such as Hindi as well and many more languages. We
demonstrate the language free aspects by showing a gain of 12% for two review
datasets over earlier results, and also release a new larger dataset for future
testing (Singh,2015).
|
Efficient Deep Feature Learning and Extraction via StochasticNets | Deep neural networks are a powerful tool for feature learning and extraction
given their ability to model high-level abstractions in highly complex data.
One area worth exploring in feature learning and extraction using deep neural
networks is efficient neural connectivity formation for faster feature learning
and extraction. Motivated by findings of stochastic synaptic connectivity
formation in the brain as well as the brain's uncanny ability to efficiently
represent information, we propose the efficient learning and extraction of
features via StochasticNets, where sparsely-connected deep neural networks can
be formed via stochastic connectivity between neurons. To evaluate the
feasibility of such a deep neural network architecture for feature learning and
extraction, we train deep convolutional StochasticNets to learn abstract
features using the CIFAR-10 dataset, and extract the learned features from
images to perform classification on the SVHN and STL-10 datasets. Experimental
results show that features learned using deep convolutional StochasticNets,
with fewer neural connections than conventional deep convolutional neural
networks, can allow for better or comparable classification accuracy than
conventional deep neural networks: relative test error decrease of ~4.5% for
classification on the STL-10 dataset and ~1% for classification on the SVHN
dataset. Furthermore, it was shown that the deep features extracted using deep
convolutional StochasticNets can provide comparable classification accuracy
even when only 10% of the training data is used for feature learning. Finally,
it was also shown that significant gains in feature extraction speed can be
achieved in embedded applications using StochasticNets. As such, StochasticNets
allow for faster feature learning and extraction performance while facilitate
for better or comparable accuracy performances.
|
Active Sampler: Light-weight Accelerator for Complex Data Analytics at
Scale | Recent years have witnessed amazing outcomes from "Big Models" trained by
"Big Data". Most popular algorithms for model training are iterative. Due to
the surging volumes of data, we can usually afford to process only a fraction
of the training data in each iteration. Typically, the data are either
uniformly sampled or sequentially accessed.
In this paper, we study how the data access pattern can affect model
training. We propose an Active Sampler algorithm, where training data with more
"learning value" to the model are sampled more frequently. The goal is to focus
training effort on valuable instances near the classification boundaries,
rather than evident cases, noisy data or outliers. We show the correctness and
optimality of Active Sampler in theory, and then develop a light-weight
vectorized implementation. Active Sampler is orthogonal to most approaches
optimizing the efficiency of large-scale data analytics, and can be applied to
most analytics models trained by stochastic gradient descent (SGD) algorithm.
Extensive experimental evaluations demonstrate that Active Sampler can speed up
the training procedure of SVM, feature selection and deep learning, for
comparable training quality by 1.6-2.2x.
|
Quantum assisted Gaussian process regression | Gaussian processes (GP) are a widely used model for regression problems in
supervised machine learning. Implementation of GP regression typically requires
$O(n^3)$ logic gates. We show that the quantum linear systems algorithm [Harrow
et al., Phys. Rev. Lett. 103, 150502 (2009)] can be applied to Gaussian process
regression (GPR), leading to an exponential reduction in computation time in
some instances. We show that even in some cases not ideally suited to the
quantum linear systems algorithm, a polynomial increase in efficiency still
occurs.
|
Active Distance-Based Clustering using K-medoids | k-medoids algorithm is a partitional, centroid-based clustering algorithm
which uses pairwise distances of data points and tries to directly decompose
the dataset with $n$ points into a set of $k$ disjoint clusters. However,
k-medoids itself requires all distances between data points that are not so
easy to get in many applications. In this paper, we introduce a new method
which requires only a small proportion of the whole set of distances and makes
an effort to estimate an upper-bound for unknown distances using the inquired
ones. This algorithm makes use of the triangle inequality to calculate an
upper-bound estimation of the unknown distances. Our method is built upon a
recursive approach to cluster objects and to choose some points actively from
each bunch of data and acquire the distances between these prominent points
from oracle. Experimental results show that the proposed method using only a
small subset of the distances can find proper clustering on many real-world and
synthetic datasets.
|
The Power of Depth for Feedforward Neural Networks | We show that there is a simple (approximately radial) function on $\reals^d$,
expressible by a small 3-layer feedforward neural networks, which cannot be
approximated by any 2-layer network, to more than a certain constant accuracy,
unless its width is exponential in the dimension. The result holds for
virtually all known activation functions, including rectified linear units,
sigmoids and thresholds, and formally demonstrates that depth -- even if
increased by 1 -- can be exponentially more valuable than width for standard
feedforward neural networks. Moreover, compared to related results in the
context of Boolean functions, our result requires fewer assumptions, and the
proof techniques and construction are very different.
|
Quantum Privacy-Preserving Data Mining | Data mining is a key technology in big data analytics and it can discover
understandable knowledge (patterns) hidden in large data sets. Association rule
is one of the most useful knowledge patterns, and a large number of algorithms
have been developed in the data mining literature to generate association rules
corresponding to different problems and situations. Privacy becomes a vital
issue when data mining is used to sensitive data sets like medical records,
commercial data sets and national security. In this Letter, we present a
quantum protocol for mining association rules on vertically partitioned
databases. The quantum protocol can improve the privacy level preserved by
known classical protocols and at the same time it can exponentially reduce the
computational complexity and communication cost.
|
L1-Regularized Distributed Optimization: A Communication-Efficient
Primal-Dual Framework | Despite the importance of sparsity in many large-scale applications, there
are few methods for distributed optimization of sparsity-inducing objectives.
In this paper, we present a communication-efficient framework for
L1-regularized optimization in the distributed environment. By viewing
classical objectives in a more general primal-dual setting, we develop a new
class of methods that can be efficiently distributed and applied to common
sparsity-inducing models, such as Lasso, sparse logistic regression, and
elastic net-regularized problems. We provide theoretical convergence guarantees
for our framework, and demonstrate its efficiency and flexibility with a
thorough experimental comparison on Amazon EC2. Our proposed framework yields
speedups of up to 50x as compared to current state-of-the-art methods for
distributed L1-regularized optimization.
|
Tracking Idea Flows between Social Groups | In many applications, ideas that are described by a set of words often flow
between different groups. To facilitate users in analyzing the flow, we present
a method to model the flow behaviors that aims at identifying the lead-lag
relationships between word clusters of different user groups. In particular, an
improved Bayesian conditional cointegration based on dynamic time warping is
employed to learn links between words in different groups. A tensor-based
technique is developed to cluster these linked words into different clusters
(ideas) and track the flow of ideas. The main feature of the tensor
representation is that we introduce two additional dimensions to represent both
time and lead-lag relationships. Experiments on both synthetic and real
datasets show that our method is more effective than methods based on
traditional clustering techniques and achieves better accuracy. A case study
was conducted to demonstrate the usefulness of our method in helping users
understand the flow of ideas between different user groups on social media
|
Distributed Optimization with Arbitrary Local Solvers | With the growth of data and necessity for distributed optimization methods,
solvers that work well on a single machine must be re-designed to leverage
distributed computation. Recent work in this area has been limited by focusing
heavily on developing highly specific methods for the distributed environment.
These special-purpose methods are often unable to fully leverage the
competitive performance of their well-tuned and customized single machine
counterparts. Further, they are unable to easily integrate improvements that
continue to be made to single machine methods. To this end, we present a
framework for distributed optimization that both allows the flexibility of
arbitrary solvers to be used on each (single) machine locally, and yet
maintains competitive performance against other state-of-the-art
special-purpose distributed methods. We give strong primal-dual convergence
rate guarantees for our framework that hold for arbitrary local solvers. We
demonstrate the impact of local solver selection both theoretically and in an
extensive experimental comparison. Finally, we provide thorough implementation
details for our framework, highlighting areas for practical performance gains.
|
Big Data Scaling through Metric Mapping: Exploiting the Remarkable
Simplicity of Very High Dimensional Spaces using Correspondence Analysis | We present new findings in regard to data analysis in very high dimensional
spaces. We use dimensionalities up to around one million. A particular benefit
of Correspondence Analysis is its suitability for carrying out an orthonormal
mapping, or scaling, of power law distributed data. Power law distributed data
are found in many domains. Correspondence factor analysis provides a latent
semantic or principal axes mapping. Our experiments use data from digital
chemistry and finance, and other statistically generated data.
|
True Online Temporal-Difference Learning | The temporal-difference methods TD($\lambda$) and Sarsa($\lambda$) form a
core part of modern reinforcement learning. Their appeal comes from their good
performance, low computational cost, and their simple interpretation, given by
their forward view. Recently, new versions of these methods were introduced,
called true online TD($\lambda$) and true online Sarsa($\lambda$), respectively
(van Seijen & Sutton, 2014). These new versions maintain an exact equivalence
with the forward view at all times, whereas the traditional versions only
approximate it for small step-sizes. We hypothesize that these true online
methods not only have better theoretical properties, but also dominate the
regular methods empirically. In this article, we put this hypothesis to the
test by performing an extensive empirical comparison. Specifically, we compare
the performance of true online TD($\lambda$)/Sarsa($\lambda$) with regular
TD($\lambda$)/Sarsa($\lambda$) on random MRPs, a real-world myoelectric
prosthetic arm, and a domain from the Arcade Learning Environment. We use
linear function approximation with tabular, binary, and non-binary features.
Our results suggest that the true online methods indeed dominate the regular
methods. Across all domains/representations the learning speed of the true
online methods are often better, but never worse than that of the regular
methods. An additional advantage is that no choice between traces has to be
made for the true online methods. Besides the empirical results, we provide an
in-depth analysis of the theory behind true online temporal-difference
learning. In addition, we show that new true online temporal-difference methods
can be derived by making changes to the online forward view and then rewriting
the update equations.
|
Stack Exchange Tagger | The goal of our project is to develop an accurate tagger for questions posted
on Stack Exchange. Our problem is an instance of the more general problem of
developing accurate classifiers for large scale text datasets. We are tackling
the multilabel classification problem where each item (in this case, question)
can belong to multiple classes (in this case, tags). We are predicting the tags
(or keywords) for a particular Stack Exchange post given only the question text
and the title of the post. In the process, we compare the performance of
Support Vector Classification (SVC) for different kernel functions, loss
function, etc. We found linear SVC with Crammer Singer technique produces best
results.
|
Policy Gradient Methods for Off-policy Control | Off-policy learning refers to the problem of learning the value function of a
way of behaving, or policy, while following a different policy. Gradient-based
off-policy learning algorithms, such as GTD and TDC/GQ, converge even when
using function approximation and incremental updates. However, they have been
developed for the case of a fixed behavior policy. In control problems, one
would like to adapt the behavior policy over time to become more greedy with
respect to the existing value function. In this paper, we present the first
gradient-based learning algorithms for this problem, which rely on the
framework of policy gradient in order to modify the behavior policy. We present
derivations of the algorithms, a convergence theorem, and empirical evidence
showing that they compare favorably to existing approaches.
|
Fighting Bandits with a New Kind of Smoothness | We define a novel family of algorithms for the adversarial multi-armed bandit
problem, and provide a simple analysis technique based on convex smoothing. We
prove two main results. First, we show that regularization via the
\emph{Tsallis entropy}, which includes EXP3 as a special case, achieves the
$\Theta(\sqrt{TN})$ minimax regret. Second, we show that a wide class of
perturbation methods achieve a near-optimal regret as low as $O(\sqrt{TN \log
N})$ if the perturbation distribution has a bounded hazard rate. For example,
the Gumbel, Weibull, Frechet, Pareto, and Gamma distributions all satisfy this
key property.
|
Preconditioned Stochastic Gradient Descent | Stochastic gradient descent (SGD) still is the workhorse for many practical
problems. However, it converges slow, and can be difficult to tune. It is
possible to precondition SGD to accelerate its convergence remarkably. But many
attempts in this direction either aim at solving specialized problems, or
result in significantly more complicated methods than SGD. This paper proposes
a new method to estimate a preconditioner such that the amplitudes of
perturbations of preconditioned stochastic gradient match that of the
perturbations of parameters to be optimized in a way comparable to Newton
method for deterministic optimization. Unlike the preconditioners based on
secant equation fitting as done in deterministic quasi-Newton methods, which
assume positive definite Hessian and approximate its inverse, the new
preconditioner works equally well for both convex and non-convex optimizations
with exact or noisy gradients. When stochastic gradient is used, it can
naturally damp the gradient noise to stabilize SGD. Efficient preconditioner
estimation methods are developed, and with reasonable simplifications, they are
applicable to large scaled problems. Experimental results demonstrate that
equipped with the new preconditioner, without any tuning effort, preconditioned
SGD can efficiently solve many challenging problems like the training of a deep
neural network or a recurrent neural network requiring extremely long term
memories.
|
Small-footprint Deep Neural Networks with Highway Connections for Speech
Recognition | For speech recognition, deep neural networks (DNNs) have significantly
improved the recognition accuracy in most of benchmark datasets and application
domains. However, compared to the conventional Gaussian mixture models,
DNN-based acoustic models usually have much larger number of model parameters,
making it challenging for their applications in resource constrained platforms,
e.g., mobile devices. In this paper, we study the application of the recently
proposed highway network to train small-footprint DNNs, which are {\it thinner}
and {\it deeper}, and have significantly smaller number of model parameters
compared to conventional DNNs. We investigated this approach on the AMI meeting
speech transcription corpus which has around 70 hours of audio data. The
highway neural networks constantly outperformed their plain DNN counterparts,
and the number of model parameters can be reduced significantly without
sacrificing the recognition accuracy.
|
Origami: A 803 GOp/s/W Convolutional Network Accelerator | An ever increasing number of computer vision and image/video processing
challenges are being approached using deep convolutional neural networks,
obtaining state-of-the-art results in object recognition and detection,
semantic segmentation, action recognition, optical flow and superresolution.
Hardware acceleration of these algorithms is essential to adopt these
improvements in embedded and mobile computer vision systems. We present a new
architecture, design and implementation as well as the first reported silicon
measurements of such an accelerator, outperforming previous work in terms of
power-, area- and I/O-efficiency. The manufactured device provides up to 196
GOp/s on 3.09 mm^2 of silicon in UMC 65nm technology and can achieve a power
efficiency of 803 GOp/s/W. The massively reduced bandwidth requirements make it
the first architecture scalable to TOp/s performance.
|
Automatic Incident Classification for Big Traffic Data by Adaptive
Boosting SVM | Modern cities experience heavy traffic flows and congestions regularly across
space and time. Monitoring traffic situations becomes an important challenge
for the Traffic Control and Surveillance Systems (TCSS). In advanced TCSS, it
is helpful to automatically detect and classify different traffic incidents
such as severity of congestion, abnormal driving pattern, abrupt or illegal
stop on road, etc. Although most TCSS are equipped with basic incident
detection algorithms, they are however crude to be really useful as an
automated tool for further classification. In literature, there is a lack of
research for Automated Incident Classification (AIC). Therefore, a novel AIC
method is proposed in this paper to tackle such challenges. In the proposed
method, traffic signals are firstly extracted from captured videos and
converted as spatial-temporal (ST) signals. Based on the characteristics of the
ST signals, a set of realistic simulation data are generated to construct an
extended big traffic database to cover a variety of traffic situations. Next, a
Mean-Shift filter is introduced to suppress the effect of noise and extract
significant features from the ST signals. The extracted features are then
associated with various types of traffic data: one normal type (inliers) and
multiple abnormal types (outliers). For the classification, an adaptive
boosting classifier is trained to detect outliers in traffic data
automatically. Further, a Support Vector Machine (SVM) based method is adopted
to train the model for identifying the categories of outliers. In short, this
hybrid approach is called an Adaptive Boosting Support Vector Machines (AB-SVM)
method. Experimental results show that the proposed AB-SVM method achieves a
satisfied result with more than 92% classification accuracy on average.
|
We Are Humor Beings: Understanding and Predicting Visual Humor | Humor is an integral part of human lives. Despite being tremendously
impactful, it is perhaps surprising that we do not have a detailed
understanding of humor yet. As interactions between humans and AI systems
increase, it is imperative that these systems are taught to understand
subtleties of human expressions such as humor. In this work, we are interested
in the question - what content in a scene causes it to be funny? As a first
step towards understanding visual humor, we analyze the humor manifested in
abstract scenes and design computational models for them. We collect two
datasets of abstract scenes that facilitate the study of humor at both the
scene-level and the object-level. We analyze the funny scenes and explore the
different types of humor depicted in them via human studies. We model two tasks
that we believe demonstrate an understanding of some aspects of visual humor.
The tasks involve predicting the funniness of a scene and altering the
funniness of a scene. We show that our models perform well quantitatively, and
qualitatively through human studies. Our datasets are publicly available.
|
Near-Optimal Bounds for Binary Embeddings of Arbitrary Sets | We study embedding a subset $K$ of the unit sphere to the Hamming cube
$\{-1,+1\}^m$. We characterize the tradeoff between distortion and sample
complexity $m$ in terms of the Gaussian width $\omega(K)$ of the set. For
subspaces and several structured sets we show that Gaussian maps provide the
optimal tradeoff $m\sim \delta^{-2}\omega^2(K)$, in particular for $\delta$
distortion one needs $m\approx\delta^{-2}{d}$ where $d$ is the subspace
dimension. For general sets, we provide sharp characterizations which reduces
to $m\approx{\delta^{-4}}{\omega^2(K)}$ after simplification. We provide
improved results for local embedding of points that are in close proximity of
each other which is related to locality sensitive hashing. We also discuss
faster binary embedding where one takes advantage of an initial sketching
procedure based on Fast Johnson-Lindenstauss Transform. Finally, we list
several numerical observations and discuss open problems.
|
Memory-based control with recurrent neural networks | Partially observed control problems are a challenging aspect of reinforcement
learning. We extend two related, model-free algorithms for continuous control
-- deterministic policy gradient and stochastic value gradient -- to solve
partially observed domains using recurrent neural networks trained with
backpropagation through time.
We demonstrate that this approach, coupled with long-short term memory is
able to solve a variety of physical control problems exhibiting an assortment
of memory requirements. These include the short-term integration of information
from noisy sensors and the identification of system parameters, as well as
long-term memory problems that require preserving information over many time
steps. We also demonstrate success on a combined exploration and memory problem
in the form of a simplified version of the well-known Morris water maze task.
Finally, we show that our approach can deal with high-dimensional observations
by learning directly from pixels.
We find that recurrent deterministic and stochastic policies are able to
learn similarly good solutions to these tasks, including the water maze where
the agent must learn effective search strategies.
|
Semisupervised Autoencoder for Sentiment Analysis | In this paper, we investigate the usage of autoencoders in modeling textual
data. Traditional autoencoders suffer from at least two aspects: scalability
with the high dimensionality of vocabulary size and dealing with
task-irrelevant words. We address this problem by introducing supervision via
the loss function of autoencoders. In particular, we first train a linear
classifier on the labeled data, then define a loss for the autoencoder with the
weights learned from the linear classifier. To reduce the bias brought by one
single classifier, we define a posterior probability distribution on the
weights of the classifier, and derive the marginalized loss of the autoencoder
with Laplace approximation. We show that our choice of loss function can be
rationalized from the perspective of Bregman Divergence, which justifies the
soundness of our model. We evaluate the effectiveness of our model on six
sentiment analysis datasets, and show that our model significantly outperforms
all the competing methods with respect to classification accuracy. We also show
that our model is able to take advantage of unlabeled dataset and get improved
performance. We further show that our model successfully learns highly
discriminative feature maps, which explains its superior performance.
|
\"Uber die Klassifizierung von Knoten in dynamischen Netzwerken mit
Inhalt | This paper explains the DYCOS-Algorithm as it was introduced in by Aggarwal
and Li in 2011. It operates on graphs whichs nodes are partially labeled and
automatically adds missing labels to nodes. To do so, the DYCOS algorithm makes
use of the structure of the graph as well as content which is assigned to the
node. Aggarwal and Li measured in an experimental analysis that DYCOS adds the
missing labels to a Graph with 19396 nodes of which 14814 are labeled and
another Graph with 806635 nodes of which 18999 are labeld on one core of an
Intel Xeon 2.5 GHz CPU with 32 G RAM within less than a minute. Additionally,
extensions of the DYCOS algorithm are proposed.
-----
In dieser Arbeit wird der DYCOS-Algorithmus, wie er 2011 von Aggarwal und Li
vorgestellt wurde, erkl\"art. Er arbeitet auf Graphen, deren Knoten teilweise
mit Beschriftungen versehen sind und erg\"anzt automatisch Beschriftungen f\"ur
Knoten, die bisher noch keine Beschriftung haben. Dieser Vorgang wird
"Klassifizierung" genannt. Dazu verwendet er die Struktur des Graphen sowie
textuelle Informationen, die den Knoten zugeordnet sind. Die von Aggarwal und
Li beschriebene experimentelle Analyse ergab, dass er auch auf dynamischen
Graphen mit 19396 bzw. 806635 Knoten, von denen nur 14814 bzw. 18999
beschriftet waren, innerhalb von weniger als einer Minute auf einem Kern einer
Intel Xeon 2.5 GHz CPU mit 32 G RAM ausgef\"uhrt werden kann. Zus\"atzlich wird
die Ver\"offentlichung von Aggarwal und Li kritisch er\"ortert und und es
werden m\"ogliche Erweiterungen des DYCOS-Algorithmus vorgeschlagen.
|
Dropout Training of Matrix Factorization and Autoencoder for Link
Prediction in Sparse Graphs | Matrix factorization (MF) and Autoencoder (AE) are among the most successful
approaches of unsupervised learning. While MF based models have been
extensively exploited in the graph modeling and link prediction literature, the
AE family has not gained much attention. In this paper we investigate both MF
and AE's application to the link prediction problem in sparse graphs. We show
the connection between AE and MF from the perspective of multiview learning,
and further propose MF+AE: a model training MF and AE jointly with shared
parameters. We apply dropout to training both the MF and AE parts, and show
that it can significantly prevent overfitting by acting as an adaptive
regularization. We conduct experiments on six real world sparse graph datasets,
and show that MF+AE consistently outperforms the competing methods, especially
on datasets that demonstrate strong non-cohesive structures.
|
On non-iterative training of a neural classifier | Recently an algorithm, was discovered, which separates points in n-dimension
by planes in such a manner that no two points are left un-separated by at least
one plane{[}1-3{]}. By using this new algorithm we show that there are two ways
of classification by a neural network, for a large dimension feature space,
both of which are non-iterative and deterministic. To demonstrate the power of
both these methods we apply them exhaustively to the classical pattern
recognition problem: The Fisher-Anderson's, IRIS flower data set and present
the results.
It is expected these methods will now be widely used for the training of
neural networks for Deep Learning not only because of their non-iterative and
deterministic nature but also because of their efficiency and speed and will
supersede other classification methods which are iterative in nature and rely
on error minimization.
|
Relaxed Linearized Algorithms for Faster X-Ray CT Image Reconstruction | Statistical image reconstruction (SIR) methods are studied extensively for
X-ray computed tomography (CT) due to the potential of acquiring CT scans with
reduced X-ray dose while maintaining image quality. However, the longer
reconstruction time of SIR methods hinders their use in X-ray CT in practice.
To accelerate statistical methods, many optimization techniques have been
investigated. Over-relaxation is a common technique to speed up convergence of
iterative algorithms. For instance, using a relaxation parameter that is close
to two in alternating direction method of multipliers (ADMM) has been shown to
speed up convergence significantly. This paper proposes a relaxed linearized
augmented Lagrangian (AL) method that shows theoretical faster convergence rate
with over-relaxation and applies the proposed relaxed linearized AL method to
X-ray CT image reconstruction problems. Experimental results with both
simulated and real CT scan data show that the proposed relaxed algorithm (with
ordered-subsets [OS] acceleration) is about twice as fast as the existing
unrelaxed fast algorithms, with negligible computation and memory overhead.
|
Learning optimal nonlinearities for iterative thresholding algorithms | Iterative shrinkage/thresholding algorithm (ISTA) is a well-studied method
for finding sparse solutions to ill-posed inverse problems. In this letter, we
present a data-driven scheme for learning optimal thresholding functions for
ISTA. The proposed scheme is obtained by relating iterations of ISTA to layers
of a simple deep neural network (DNN) and developing a corresponding error
backpropagation algorithm that allows to fine-tune the thresholding functions.
Simulations on sparse statistical signals illustrate potential gains in
estimation quality due to the proposed data adaptive ISTA.
|
From One Point to A Manifold: Knowledge Graph Embedding For Precise Link
Prediction | Knowledge graph embedding aims at offering a numerical knowledge
representation paradigm by transforming the entities and relations into
continuous vector space. However, existing methods could not characterize the
knowledge graph in a fine degree to make a precise prediction. There are two
reasons: being an ill-posed algebraic system and applying an overstrict
geometric form. As precise prediction is critical, we propose an manifold-based
embedding principle (\textbf{ManifoldE}) which could be treated as a well-posed
algebraic system that expands the position of golden triples from one point in
current models to a manifold in ours. Extensive experiments show that the
proposed models achieve substantial improvements against the state-of-the-art
baselines especially for the precise prediction task, and yet maintain high
efficiency.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.