title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
A Generative Model for Deep Convolutional Learning | A generative model is developed for deep (multi-layered) convolutional
dictionary learning. A novel probabilistic pooling operation is integrated into
the deep model, yielding efficient bottom-up (pretraining) and top-down
(refinement) probabilistic learning. Experimental results demonstrate powerful
capabilities of the model to learn multi-layer features from images, and
excellent classification results are obtained on the MNIST and Caltech 101
datasets.
|
Faster Algorithms for Testing under Conditional Sampling | There has been considerable recent interest in distribution-tests whose
run-time and sample requirements are sublinear in the domain-size $k$. We study
two of the most important tests under the conditional-sampling model where each
query specifies a subset $S$ of the domain, and the response is a sample drawn
from $S$ according to the underlying distribution.
For identity testing, which asks whether the underlying distribution equals a
specific given distribution or $\epsilon$-differs from it, we reduce the known
time and sample complexities from $\tilde{\mathcal{O}}(\epsilon^{-4})$ to
$\tilde{\mathcal{O}}(\epsilon^{-2})$, thereby matching the information
theoretic lower bound. For closeness testing, which asks whether two
distributions underlying observed data sets are equal or different, we reduce
existing complexity from $\tilde{\mathcal{O}}(\epsilon^{-4} \log^5 k)$ to an
even sub-logarithmic $\tilde{\mathcal{O}}(\epsilon^{-5} \log \log k)$ thus
providing a better bound to an open problem in Bertinoro Workshop on Sublinear
Algorithms [Fisher, 2004].
|
Actively Learning to Attract Followers on Twitter | Twitter, a popular social network, presents great opportunities for on-line
machine learning research. However, previous research has focused almost
entirely on learning from passively collected data. We study the problem of
learning to acquire followers through normative user behavior, as opposed to
the mass following policies applied by many bots. We formalize the problem as a
contextual bandit problem, in which we consider retweeting content to be the
action chosen and each tweet (content) is accompanied by context. We design
reward signals based on the change in followers. The result of our month long
experiment with 60 agents suggests that (1) aggregating experience across
agents can adversely impact prediction accuracy and (2) the Twitter community's
response to different actions is non-stationary. Our findings suggest that
actively learning on-line can provide deeper insights about how to attract
followers than machine learning over passively collected data alone.
|
Caffe con Troll: Shallow Ideas to Speed Up Deep Learning | We present Caffe con Troll (CcT), a fully compatible end-to-end version of
the popular framework Caffe with rebuilt internals. We built CcT to examine the
performance characteristics of training and deploying general-purpose
convolutional neural networks across different hardware architectures. We find
that, by employing standard batching optimizations for CPU training, we achieve
a 4.5x throughput improvement over Caffe on popular networks like CaffeNet.
Moreover, with these improvements, the end-to-end training time for CNNs is
directly proportional to the FLOPS delivered by the CPU, which enables us to
efficiently train hybrid CPU-GPU systems for CNNs.
|
Non-Uniform Stochastic Average Gradient Method for Training Conditional
Random Fields | We apply stochastic average gradient (SAG) algorithms for training
conditional random fields (CRFs). We describe a practical implementation that
uses structure in the CRF gradient to reduce the memory requirement of this
linearly-convergent stochastic gradient method, propose a non-uniform sampling
scheme that substantially improves practical performance, and analyze the rate
of convergence of the SAGA variant under non-uniform sampling. Our experimental
results reveal that our method often significantly outperforms existing methods
in terms of the training objective, and performs as well or better than
optimally-tuned stochastic gradient methods in terms of test error.
|
Mini-Batch Semi-Stochastic Gradient Descent in the Proximal Setting | We propose mS2GD: a method incorporating a mini-batching scheme for improving
the theoretical complexity and practical performance of semi-stochastic
gradient descent (S2GD). We consider the problem of minimizing a strongly
convex function represented as the sum of an average of a large number of
smooth convex functions, and a simple nonsmooth convex regularizer. Our method
first performs a deterministic step (computation of the gradient of the
objective function at the starting point), followed by a large number of
stochastic steps. The process is repeated a few times with the last iterate
becoming the new starting point. The novelty of our method is in introduction
of mini-batching into the computation of stochastic steps. In each step,
instead of choosing a single function, we sample $b$ functions, compute their
gradients, and compute the direction based on this. We analyze the complexity
of the method and show that it benefits from two speedup effects. First, we
prove that as long as $b$ is below a certain threshold, we can reach any
predefined accuracy with less overall work than without mini-batching. Second,
our mini-batching scheme admits a simple parallel implementation, and hence is
suitable for further acceleration by parallelization.
|
The Nataf-Beta Random Field Classifier: An Extension of the Beta
Conjugate Prior to Classification Problems | This paper presents the Nataf-Beta Random Field Classifier, a discriminative
approach that extends the applicability of the Beta conjugate prior to
classification problems. The approach's key feature is to model the probability
of a class conditional on attribute values as a random field whose marginals
are Beta distributed, and where the parameters of marginals are themselves
described by random fields. Although the classification accuracy of the
approach proposed does not statistically outperform the best accuracies
reported in the literature, it ranks among the top tier for the six benchmark
datasets tested. The Nataf-Beta Random Field Classifier is suited as a general
purpose classification approach for real-continuous and real-integer attribute
value problems.
|
Testing Closeness With Unequal Sized Samples | We consider the problem of closeness testing for two discrete distributions
in the practically relevant setting of \emph{unequal} sized samples drawn from
each of them. Specifically, given a target error parameter $\varepsilon > 0$,
$m_1$ independent draws from an unknown distribution $p,$ and $m_2$ draws from
an unknown distribution $q$, we describe a test for distinguishing the case
that $p=q$ from the case that $||p-q||_1 \geq \varepsilon$. If $p$ and $q$ are
supported on at most $n$ elements, then our test is successful with high
probability provided $m_1\geq n^{2/3}/\varepsilon^{4/3}$ and $m_2 =
\Omega(\max\{\frac{n}{\sqrt m_1\varepsilon^2}, \frac{\sqrt
n}{\varepsilon^2}\});$ we show that this tradeoff is optimal throughout this
range, to constant factors. These results extend the recent work of Chan et al.
who established the sample complexity when the two samples have equal sizes,
and tightens the results of Acharya et al. by polynomials factors in both $n$
and $\varepsilon$. As a consequence, we obtain an algorithm for estimating the
mixing time of a Markov chain on $n$ states up to a $\log n$ factor that uses
$\tilde{O}(n^{3/2} \tau_{mix})$ queries to a "next node" oracle, improving upon
the $\tilde{O}(n^{5/3}\tau_{mix})$ query algorithm of Batu et al. Finally, we
note that the core of our testing algorithm is a relatively simple statistic
that seems to perform well in practice, both on synthetic data and on natural
language data.
|
Performance Evaluation of Machine Learning Algorithms in Post-operative
Life Expectancy in the Lung Cancer Patients | The nature of clinical data makes it difficult to quickly select, tune and
apply machine learning algorithms to clinical prognosis. As a result, a lot of
time is spent searching for the most appropriate machine learning algorithms
applicable in clinical prognosis that contains either binary-valued or
multi-valued attributes. The study set out to identify and evaluate the
performance of machine learning classification schemes applied in clinical
prognosis of post-operative life expectancy in the lung cancer patients.
Multilayer Perceptron, J48, and the Naive Bayes algorithms were used to train
and test models on Thoracic Surgery datasets obtained from the University of
California Irvine machine learning repository. Stratified 10-fold
cross-validation was used to evaluate baseline performance accuracy of the
classifiers. The comparative analysis shows that multilayer perceptron
performed best with classification accuracy of 82.3%, J48 came out second with
classification accuracy of 81.8%, and Naive Bayes came out the worst with
classification accuracy of 74.4%. The quality and outcome of the chosen machine
learning algorithms depends on the ingenuity of the clinical miner.
|
Deep Karaoke: Extracting Vocals from Musical Mixtures Using a
Convolutional Deep Neural Network | Identification and extraction of singing voice from within musical mixtures
is a key challenge in source separation and machine audition. Recently, deep
neural networks (DNN) have been used to estimate 'ideal' binary masks for
carefully controlled cocktail party speech separation problems. However, it is
not yet known whether these methods are capable of generalizing to the
discrimination of voice and non-voice in the context of musical mixtures. Here,
we trained a convolutional DNN (of around a billion parameters) to provide
probabilistic estimates of the ideal binary mask for separation of vocal sounds
from real-world musical mixtures. We contrast our DNN results with more
traditional linear methods. Our approach may be useful for automatic removal of
vocal sounds from musical mixtures for 'karaoke' type applications.
|
Unsupervised Dependency Parsing: Let's Use Supervised Parsers | We present a self-training approach to unsupervised dependency parsing that
reuses existing supervised and unsupervised parsing algorithms. Our approach,
called `iterated reranking' (IR), starts with dependency trees generated by an
unsupervised parser, and iteratively improves these trees using the richer
probability models used in supervised parsing that are in turn trained on these
trees. Our system achieves 1.8% accuracy higher than the state-of-the-part
parser of Spitkovsky et al. (2013) on the WSJ corpus.
|
Local, Private, Efficient Protocols for Succinct Histograms | We give efficient protocols and matching accuracy lower bounds for frequency
estimation in the local model for differential privacy. In this model,
individual users randomize their data themselves, sending differentially
private reports to an untrusted server that aggregates them.
We study protocols that produce a succinct histogram representation of the
data. A succinct histogram is a list of the most frequent items in the data
(often called "heavy hitters") along with estimates of their frequencies; the
frequency of all other items is implicitly estimated as 0.
If there are $n$ users whose items come from a universe of size $d$, our
protocols run in time polynomial in $n$ and $\log(d)$. With high probability,
they estimate the accuracy of every item up to error
$O\left(\sqrt{\log(d)/(\epsilon^2n)}\right)$ where $\epsilon$ is the privacy
parameter. Moreover, we show that this much error is necessary, regardless of
computational efficiency, and even for the simple setting where only one item
appears with significant frequency in the data set.
Previous protocols (Mishra and Sandler, 2006; Hsu, Khanna and Roth, 2012) for
this task either ran in time $\Omega(d)$ or had much worse error (about
$\sqrt[6]{\log(d)/(\epsilon^2n)}$), and the only known lower bound on error was
$\Omega(1/\sqrt{n})$.
We also adapt a result of McGregor et al (2010) to the local setting. In a
model with public coins, we show that each user need only send 1 bit to the
server. For all known local protocols (including ours), the transformation
preserves computational efficiency.
|
Fast optimization of Multithreshold Entropy Linear Classifier | Multithreshold Entropy Linear Classifier (MELC) is a density based model
which searches for a linear projection maximizing the Cauchy-Schwarz Divergence
of dataset kernel density estimation. Despite its good empirical results, one
of its drawbacks is the optimization speed. In this paper we analyze how one
can speed it up through solving an approximate problem. We analyze two methods,
both similar to the approximate solutions of the Kernel Density Estimation
querying and provide adaptive schemes for selecting a crucial parameters based
on user-specified acceptable error. Furthermore we show how one can exploit
well known conjugate gradients and L-BFGS optimizers despite the fact that the
original optimization problem should be solved on the sphere. All above methods
and modifications are tested on 10 real life datasets from UCI repository to
confirm their practical usability.
|
On the consistency of Multithreshold Entropy Linear Classifier | Multithreshold Entropy Linear Classifier (MELC) is a recent classifier idea
which employs information theoretic concept in order to create a multithreshold
maximum margin model. In this paper we analyze its consistency over
multithreshold linear models and show that its objective function upper bounds
the amount of misclassified points in a similar manner like hinge loss does in
support vector machines. For further confirmation we also conduct some
numerical experiments on five datasets.
|
Online Inference for Relation Extraction with a Reduced Feature Set | Access to web-scale corpora is gradually bringing robust automatic knowledge
base creation and extension within reach. To exploit these large
unannotated---and extremely difficult to annotate---corpora, unsupervised
machine learning methods are required. Probabilistic models of text have
recently found some success as such a tool, but scalability remains an obstacle
in their application, with standard approaches relying on sampling schemes that
are known to be difficult to scale. In this report, we therefore present an
empirical assessment of the sublinear time sparse stochastic variational
inference (SSVI) scheme applied to RelLDA. We demonstrate that online inference
leads to relatively strong qualitative results but also identify some of its
pathologies---and those of the model---which will need to be overcome if SSVI
is to be used for large-scale relation extraction.
|
Compressing Neural Networks with the Hashing Trick | As deep nets are increasingly used in applications suited for mobile devices,
a fundamental dilemma becomes apparent: the trend in deep learning is to grow
models to absorb ever-increasing data set sizes; however mobile devices are
designed with very little memory and cannot store such large models. We present
a novel network architecture, HashedNets, that exploits inherent redundancy in
neural networks to achieve drastic reductions in model sizes. HashedNets uses a
low-cost hash function to randomly group connection weights into hash buckets,
and all connections within the same hash bucket share a single parameter value.
These parameters are tuned to adjust to the HashedNets weight sharing
architecture with standard backprop during training. Our hashing procedure
introduces no additional memory overhead, and we demonstrate on several
benchmark data sets that HashedNets shrink the storage requirements of neural
networks substantially while mostly preserving generalization performance.
|
Exploring Bayesian Models for Multi-level Clustering of Hierarchically
Grouped Sequential Data | A wide range of Bayesian models have been proposed for data that is divided
hierarchically into groups. These models aim to cluster the data at different
levels of grouping, by assigning a mixture component to each datapoint, and a
mixture distribution to each group. Multi-level clustering is facilitated by
the sharing of these components and distributions by the groups. In this paper,
we introduce the concept of Degree of Sharing (DoS) for the mixture components
and distributions, with an aim to analyze and classify various existing models.
Next we introduce a generalized hierarchical Bayesian model, of which the
existing models can be shown to be special cases. Unlike most of these models,
our model takes into account the sequential nature of the data, and various
other temporal structures at different levels while assigning mixture
components and distributions. We show one specialization of this model aimed at
hierarchical segmentation of news transcripts, and present a Gibbs Sampling
based inference algorithm for it. We also show experimentally that the proposed
model outperforms existing models for the same task.
|
F-SVM: Combination of Feature Transformation and SVM Learning via Convex
Relaxation | The generalization error bound of support vector machine (SVM) depends on the
ratio of radius and margin, while standard SVM only considers the maximization
of the margin but ignores the minimization of the radius. Several approaches
have been proposed to integrate radius and margin for joint learning of feature
transformation and SVM classifier. However, most of them either require the
form of the transformation matrix to be diagonal, or are non-convex and
computationally expensive. In this paper, we suggest a novel approximation for
the radius of minimum enclosing ball (MEB) in feature space, and then propose a
convex radius-margin based SVM model for joint learning of feature
transformation and SVM classifier, i.e., F-SVM. An alternating minimization
method is adopted to solve the F-SVM model, where the feature transformation is
updatedvia gradient descent and the classifier is updated by employing the
existing SVM solver. By incorporating with kernel principal component analysis,
F-SVM is further extended for joint learning of nonlinear transformation and
classifier. Experimental results on the UCI machine learning datasets and the
LFW face datasets show that F-SVM outperforms the standard SVM and the existing
radius-margin based SVMs, e.g., RMM, R-SVM+ and R-SVM+{\mu}.
|
Nonparametric Nearest Neighbor Random Process Clustering | We consider the problem of clustering noisy finite-length observations of
stationary ergodic random processes according to their nonparametric generative
models without prior knowledge of the model statistics and the number of
generative models. Two algorithms, both using the L1-distance between estimated
power spectral densities (PSDs) as a measure of dissimilarity, are analyzed.
The first algorithm, termed nearest neighbor process clustering (NNPC), to the
best of our knowledge, is new and relies on partitioning the nearest neighbor
graph of the observations via spectral clustering. The second algorithm, simply
referred to as k-means (KM), consists of a single k-means iteration with
farthest point initialization and was considered before in the literature,
albeit with a different measure of dissimilarity and with asymptotic
performance results only. We show that both NNPC and KM succeed with high
probability under noise and even when the generative process PSDs overlap
significantly, all provided that the observation length is sufficiently large.
Our results quantify the tradeoff between the overlap of the generative process
PSDs, the noise variance, and the observation length. Finally, we present
numerical performance results for synthetic and real data.
|
Self-Adaptive Hierarchical Sentence Model | The ability to accurately model a sentence at varying stages (e.g.,
word-phrase-sentence) plays a central role in natural language processing. As
an effort towards this goal we propose a self-adaptive hierarchical sentence
model (AdaSent). AdaSent effectively forms a hierarchy of representations from
words to phrases and then to sentences through recursive gated local
composition of adjacent segments. We design a competitive mechanism (through
gating networks) to allow the representations of the same sentence to be
engaged in a particular learning task (e.g., classification), therefore
effectively mitigating the gradient vanishing problem persistent in other
recursive models. Both qualitative and quantitative analysis shows that AdaSent
can automatically form and select the representations suitable for the task at
hand during training, yielding superior classification performance over
competitor models on 5 benchmark data sets.
|
Optimal Nudging: Solving Average-Reward Semi-Markov Decision Processes
as a Minimal Sequence of Cumulative Tasks | This paper describes a novel method to solve average-reward semi-Markov
decision processes, by reducing them to a minimal sequence of cumulative reward
problems. The usual solution methods for this type of problems update the gain
(optimal average reward) immediately after observing the result of taking an
action. The alternative introduced, optimal nudging, relies instead on setting
the gain to some fixed value, which transitorily makes the problem a
cumulative-reward task, solving it by any standard reinforcement learning
method, and only then updating the gain in a way that minimizes uncertainty in
a minmax sense. The rule for optimal gain update is derived by exploiting the
geometric features of the w-l space, a simple mapping of the space of policies.
The total number of cumulative reward tasks that need to be solved is shown to
be small. Some experiments are presented to explore the features of the
algorithm and to compare its performance with other approaches.
|
Poisson Matrix Recovery and Completion | We extend the theory of low-rank matrix recovery and completion to the case
when Poisson observations for a linear combination or a subset of the entries
of a matrix are available, which arises in various applications with count
data. We consider the usual matrix recovery formulation through maximum
likelihood with proper constraints on the matrix $M$ of size $d_1$-by-$d_2$,
and establish theoretical upper and lower bounds on the recovery error. Our
bounds for matrix completion are nearly optimal up to a factor on the order of
$\mathcal{O}(\log(d_1 d_2))$. These bounds are obtained by combing techniques
for compressed sensing for sparse vectors with Poisson noise and for analyzing
low-rank matrices, as well as adapting the arguments used for one-bit matrix
completion \cite{davenport20121} (although these two problems are different in
nature) and the adaptation requires new techniques exploiting properties of the
Poisson likelihood function and tackling the difficulties posed by the locally
sub-Gaussian characteristic of the Poisson distribution. Our results highlight
a few important distinctions of the Poisson case compared to the prior work
including having to impose a minimum signal-to-noise requirement on each
observed entry and a gap in the upper and lower bounds. We also develop a set
of efficient iterative algorithms and demonstrate their good performance on
synthetic examples and real data.
|
Decomposing Overcomplete 3rd Order Tensors using Sum-of-Squares
Algorithms | Tensor rank and low-rank tensor decompositions have many applications in
learning and complexity theory. Most known algorithms use unfoldings of tensors
and can only handle rank up to $n^{\lfloor p/2 \rfloor}$ for a $p$-th order
tensor in $\mathbb{R}^{n^p}$. Previously no efficient algorithm can decompose
3rd order tensors when the rank is super-linear in the dimension. Using ideas
from sum-of-squares hierarchy, we give the first quasi-polynomial time
algorithm that can decompose a random 3rd order tensor decomposition when the
rank is as large as $n^{3/2}/\textrm{polylog} n$.
We also give a polynomial time algorithm for certifying the injective norm of
random low rank tensors. Our tensor decomposition algorithm exploits the
relationship between injective norm and the tensor components. The proof relies
on interesting tools for decoupling random variables to prove better matrix
concentration bounds, which can be useful in other settings.
|
Distance-based species tree estimation: information-theoretic trade-off
between number of loci and sequence length under the coalescent | We consider the reconstruction of a phylogeny from multiple genes under the
multispecies coalescent. We establish a connection with the sparse signal
detection problem, where one seeks to distinguish between a distribution and a
mixture of the distribution and a sparse signal. Using this connection, we
derive an information-theoretic trade-off between the number of genes, $m$,
needed for an accurate reconstruction and the sequence length, $k$, of the
genes. Specifically, we show that to detect a branch of length $f$, one needs
$m = \Theta(1/[f^{2} \sqrt{k}])$.
|
Instance Optimal Learning | We consider the following basic learning task: given independent draws from
an unknown distribution over a discrete support, output an approximation of the
distribution that is as accurate as possible in $\ell_1$ distance (i.e. total
variation or statistical distance). Perhaps surprisingly, it is often possible
to "de-noise" the empirical distribution of the samples to return an
approximation of the true distribution that is significantly more accurate than
the empirical distribution, without relying on any prior assumptions on the
distribution. We present an instance optimal learning algorithm which optimally
performs this de-noising for every distribution for which such a de-noising is
possible. More formally, given $n$ independent draws from a distribution $p$,
our algorithm returns a labelled vector whose expected distance from $p$ is
equal to the minimum possible expected error that could be obtained by any
algorithm that knows the true unlabeled vector of probabilities of distribution
$p$ and simply needs to assign labels, up to an additive subconstant term that
is independent of $p$ and goes to zero as $n$ gets large. One conceptual
implication of this result is that for large samples, Bayesian assumptions on
the "shape" or bounds on the tail probabilities of a distribution over discrete
support are not helpful for the task of learning the distribution.
As a consequence of our techniques, we also show that given a set of $n$
samples from an arbitrary distribution, one can accurately estimate the
expected number of distinct elements that will be observed in a sample of any
size up to $n \log n$. This sort of extrapolation is practically relevant,
particularly to domains such as genomics where it is important to understand
how much more might be discovered given larger sample sizes, and we are
optimistic that our approach is practically viable.
|
Effective Discriminative Feature Selection with Non-trivial Solutions | Feature selection and feature transformation, the two main ways to reduce
dimensionality, are often presented separately. In this paper, a feature
selection method is proposed by combining the popular transformation based
dimensionality reduction method Linear Discriminant Analysis (LDA) and sparsity
regularization. We impose row sparsity on the transformation matrix of LDA
through ${\ell}_{2,1}$-norm regularization to achieve feature selection, and
the resultant formulation optimizes for selecting the most discriminative
features and removing the redundant ones simultaneously. The formulation is
extended to the ${\ell}_{2,p}$-norm regularized case: which is more likely to
offer better sparsity when $0<p<1$. Thus the formulation is a better
approximation to the feature selection problem. An efficient algorithm is
developed to solve the ${\ell}_{2,p}$-norm based optimization problem and it is
proved that the algorithm converges when $0<p\le 2$. Systematical experiments
are conducted to understand the work of the proposed method. Promising
experimental results on various types of real-world data sets demonstrate the
effectiveness of our algorithm.
|
Can FCA-based Recommender System Suggest a Proper Classifier? | The paper briefly introduces multiple classifier systems and describes a new
algorithm, which improves classification accuracy by means of recommendation of
a proper algorithm to an object classification. This recommendation is done
assuming that a classifier is likely to predict the label of the object
correctly if it has correctly classified its neighbors. The process of
assigning a classifier to each object is based on Formal Concept Analysis. We
explain the idea of the algorithm with a toy example and describe our first
experiments with real-world datasets.
|
Randomized Block Krylov Methods for Stronger and Faster Approximate
Singular Value Decomposition | Since being analyzed by Rokhlin, Szlam, and Tygert and popularized by Halko,
Martinsson, and Tropp, randomized Simultaneous Power Iteration has become the
method of choice for approximate singular value decomposition. It is more
accurate than simpler sketching algorithms, yet still converges quickly for any
matrix, independently of singular value gaps. After $\tilde{O}(1/\epsilon)$
iterations, it gives a low-rank approximation within $(1+\epsilon)$ of optimal
for spectral norm error.
We give the first provable runtime improvement on Simultaneous Iteration: a
simple randomized block Krylov method, closely related to the classic Block
Lanczos algorithm, gives the same guarantees in just
$\tilde{O}(1/\sqrt{\epsilon})$ iterations and performs substantially better
experimentally. Despite their long history, our analysis is the first of a
Krylov subspace method that does not depend on singular value gaps, which are
unreliable in practice.
Furthermore, while it is a simple accuracy benchmark, even $(1+\epsilon)$
error for spectral norm low-rank approximation does not imply that an algorithm
returns high quality principal components, a major issue for data applications.
We address this problem for the first time by showing that both Block Krylov
Iteration and a minor modification of Simultaneous Iteration give nearly
optimal PCA for any matrix. This result further justifies their strength over
non-iterative sketching methods.
Finally, we give insight beyond the worst case, justifying why both
algorithms can run much faster in practice than predicted. We clarify how
simple techniques can take advantage of common matrix properties to
significantly improve runtime.
|
Deep Convolutional Neural Networks Based on Semi-Discrete Frames | Deep convolutional neural networks have led to breakthrough results in
practical feature extraction applications. The mathematical analysis of these
networks was pioneered by Mallat, 2012. Specifically, Mallat considered
so-called scattering networks based on identical semi-discrete wavelet frames
in each network layer, and proved translation-invariance as well as deformation
stability of the resulting feature extractor. The purpose of this paper is to
develop Mallat's theory further by allowing for different and, most
importantly, general semi-discrete frames (such as, e.g., Gabor frames,
wavelets, curvelets, shearlets, ridgelets) in distinct network layers. This
allows to extract wider classes of features than point singularities resolved
by the wavelet transform. Our generalized feature extractor is proven to be
translation-invariant, and we develop deformation stability results for a
larger class of deformations than those considered by Mallat. For Mallat's
wavelet-based feature extractor, we get rid of a number of technical
conditions. The mathematical engine behind our results is continuous frame
theory, which allows us to completely detach the invariance and deformation
stability proofs from the particular algebraic structure of the underlying
frames.
|
Online Learning Algorithm for Time Series Forecasting Suitable for Low
Cost Wireless Sensor Networks Nodes | Time series forecasting is an important predictive methodology which can be
applied to a wide range of problems. Particularly, forecasting the indoor
temperature permits an improved utilization of the HVAC (Heating, Ventilating
and Air Conditioning) systems in a home and thus a better energy efficiency.
With such purpose the paper describes how to implement an Artificial Neural
Network (ANN) algorithm in a low cost system-on-chip to develop an autonomous
intelligent wireless sensor network. The present paper uses a Wireless Sensor
Networks (WSN) to monitor and forecast the indoor temperature in a smart home,
based on low resources and cost microcontroller technology as the 8051MCU. An
on-line learning approach, based on Back-Propagation (BP) algorithm for ANNs,
has been developed for real-time time series learning. It performs the model
training with every new data that arrive to the system, without saving enormous
quantities of data to create a historical database as usual, i.e., without
previous knowledge. Consequently to validate the approach a simulation study
through a Bayesian baseline model have been tested in order to compare with a
database of a real application aiming to see the performance and accuracy. The
core of the paper is a new algorithm, based on the BP one, which has been
described in detail, and the challenge was how to implement a computational
demanding algorithm in a simple architecture with very few hardware resources.
|
Temporal-Difference Networks | We introduce a generalization of temporal-difference (TD) learning to
networks of interrelated predictions. Rather than relating a single prediction
to itself at a later time, as in conventional TD methods, a TD network relates
each prediction in a set of predictions to other predictions in the set at a
later time. TD networks can represent and apply TD learning to a much wider
class of predictions than has previously been possible. Using a random-walk
example, we show that these networks can be used to learn to predict by a fixed
interval, which is not possible with conventional TD methods. Secondly, we show
that if the inter-predictive relationships are made conditional on action, then
the usual learning-efficiency advantage of TD methods over Monte Carlo
(supervised learning) methods becomes particularly pronounced. Thirdly, we
demonstrate that TD networks can learn predictive state representations that
enable exact solution of a non-Markov problem. A very broad range of
inter-predictive temporal relationships can be expressed in these networks.
Overall we argue that TD networks represent a substantial extension of the
abilities of TD methods and bring us closer to the goal of representing world
knowledge in entirely predictive, grounded terms.
|
Learning Opposites with Evolving Rules | The idea of opposition-based learning was introduced 10 years ago. Since then
a noteworthy group of researchers has used some notions of oppositeness to
improve existing optimization and learning algorithms. Among others,
evolutionary algorithms, reinforcement agents, and neural networks have been
reportedly extended into their opposition-based version to become faster and/or
more accurate. However, most works still use a simple notion of opposites,
namely linear (or type- I) opposition, that for each $x\in[a,b]$ assigns its
opposite as $\breve{x}_I=a+b-x$. This, of course, is a very naive estimate of
the actual or true (non-linear) opposite $\breve{x}_{II}$, which has been
called type-II opposite in literature. In absence of any knowledge about a
function $y=f(\mathbf{x})$ that we need to approximate, there seems to be no
alternative to the naivety of type-I opposition if one intents to utilize
oppositional concepts. But the question is if we can receive some level of
accuracy increase and time savings by using the naive opposite estimate
$\breve{x}_I$ according to all reports in literature, what would we be able to
gain, in terms of even higher accuracies and more reduction in computational
complexity, if we would generate and employ true opposites? This work
introduces an approach to approximate type-II opposites using evolving fuzzy
rules when we first perform opposition mining. We show with multiple examples
that learning true opposites is possible when we mine the opposites from the
training data to subsequently approximate $\breve{x}_{II}=f(\mathbf{x},y)$.
|
Self-Tuned Deep Super Resolution | Deep learning has been successfully applied to image super resolution (SR).
In this paper, we propose a deep joint super resolution (DJSR) model to exploit
both external and self similarities for SR. A Stacked Denoising Convolutional
Auto Encoder (SDCAE) is first pre-trained on external examples with proper data
augmentations. It is then fine-tuned with multi-scale self examples from each
input, where the reliability of self examples is explicitly taken into account.
We also enhance the model performance by sub-model training and selection. The
DJSR model is extensively evaluated and compared with state-of-the-arts, and
show noticeable performance improvements both quantitatively and perceptually
on a wide range of images.
|
Rebuilding Factorized Information Criterion: Asymptotically Accurate
Marginal Likelihood | Factorized information criterion (FIC) is a recently developed approximation
technique for the marginal log-likelihood, which provides an automatic model
selection framework for a few latent variable models (LVMs) with tractable
inference algorithms. This paper reconsiders FIC and fills theoretical gaps of
previous FIC studies. First, we reveal the core idea of FIC that allows
generalization for a broader class of LVMs, including continuous LVMs, in
contrast to previous FICs, which are applicable only to binary LVMs. Second, we
investigate the model selection mechanism of the generalized FIC. Our analysis
provides a formal justification of FIC as a model selection criterion for LVMs
and also a systematic procedure for pruning redundant latent variables that
have been removed heuristically in previous studies. Third, we provide an
interpretation of FIC as a variational free energy and uncover a few
previously-unknown their relationships. A demonstrative study on Bayesian
principal component analysis is provided and numerical experiments support our
theoretical results.
|
On the Generalization Properties of Differential Privacy | A new line of work, started with Dwork et al., studies the task of answering
statistical queries using a sample and relates the problem to the concept of
differential privacy. By the Hoeffding bound, a sample of size $O(\log
k/\alpha^2)$ suffices to answer $k$ non-adaptive queries within error $\alpha$,
where the answers are computed by evaluating the statistical queries on the
sample. This argument fails when the queries are chosen adaptively (and can
hence depend on the sample). Dwork et al. showed that if the answers are
computed with $(\epsilon,\delta)$-differential privacy then $O(\epsilon)$
accuracy is guaranteed with probability $1-O(\delta^\epsilon)$. Using the
Private Multiplicative Weights mechanism, they concluded that the sample size
can still grow polylogarithmically with the $k$.
Very recently, Bassily et al. presented an improved bound and showed that (a
variant of) the private multiplicative weights algorithm can answer $k$
adaptively chosen statistical queries using sample complexity that grows
logarithmically in $k$. However, their results no longer hold for every
differentially private algorithm, and require modifying the private
multiplicative weights algorithm in order to obtain their high probability
bounds.
We greatly simplify the results of Dwork et al. and improve on the bound by
showing that differential privacy guarantees $O(\epsilon)$ accuracy with
probability $1-O(\delta\log(1/\epsilon)/\epsilon)$. It would be tempting to
guess that an $(\epsilon,\delta)$-differentially private computation should
guarantee $O(\epsilon)$ accuracy with probability $1-O(\delta)$. However, we
show that this is not the case, and that our bound is tight (up to logarithmic
factors).
|
Learning of Behavior Trees for Autonomous Agents | Definition of an accurate system model for Automated Planner (AP) is often
impractical, especially for real-world problems. Conversely, off-the-shelf
planners fail to scale up and are domain dependent. These drawbacks are
inherited from conventional transition systems such as Finite State Machines
(FSMs) that describes the action-plan execution generated by the AP. On the
other hand, Behavior Trees (BTs) represent a valid alternative to FSMs
presenting many advantages in terms of modularity, reactiveness, scalability
and domain-independence. In this paper, we propose a model-free AP framework
using Genetic Programming (GP) to derive an optimal BT for an autonomous agent
to achieve a given goal in unknown (but fully observable) environments. We
illustrate the proposed framework using experiments conducted with an open
source benchmark Mario AI for automated generation of BTs that can play the
game character Mario to complete a certain level at various levels of
difficulty to include enemies and obstacles.
|
Normal Bandits of Unknown Means and Variances: Asymptotic Optimality,
Finite Horizon Regret Bounds, and a Solution to an Open Problem | Consider the problem of sampling sequentially from a finite number of $N \geq
2$ populations, specified by random variables $X^i_k$, $ i = 1,\ldots , N,$ and
$k = 1, 2, \ldots$; where $X^i_k$ denotes the outcome from population $i$ the
$k^{th}$ time it is sampled. It is assumed that for each fixed $i$,
$\{ X^i_k \}_{k \geq 1}$ is a sequence of i.i.d. normal random variables,
with unknown mean $\mu_i$ and unknown variance $\sigma_i^2$.
The objective is to have a policy $\pi$ for deciding from which of the $N$
populations to sample form at any time $n=1,2,\ldots$ so as to maximize the
expected sum of outcomes of $n$ samples or equivalently to minimize the regret
due to lack on information of the parameters $\mu_i$ and $\sigma_i^2$. In this
paper, we present a simple inflated sample mean (ISM) index policy that is
asymptotically optimal in the sense of Theorem 4 below. This resolves a
standing open problem from Burnetas and Katehakis (1996). Additionally, finite
horizon regret bounds are given.
|
Exploit Bounding Box Annotations for Multi-label Object Recognition | Convolutional neural networks (CNNs) have shown great performance as general
feature representations for object recognition applications. However, for
multi-label images that contain multiple objects from different categories,
scales and locations, global CNN features are not optimal. In this paper, we
incorporate local information to enhance the feature discriminative power. In
particular, we first extract object proposals from each image. With each image
treated as a bag and object proposals extracted from it treated as instances,
we transform the multi-label recognition problem into a multi-class
multi-instance learning problem. Then, in addition to extracting the typical
CNN feature representation from each proposal, we propose to make use of
ground-truth bounding box annotations (strong labels) to add another level of
local information by using nearest-neighbor relationships of local regions to
form a multi-view pipeline. The proposed multi-view multi-instance framework
utilizes both weak and strong labels effectively, and more importantly it has
the generalization ability to even boost the performance of unseen categories
by partial strong labels from other categories. Our framework is extensively
compared with state-of-the-art hand-crafted feature based methods and CNN based
methods on two multi-label benchmark datasets. The experimental results
validate the discriminative power and the generalization ability of the
proposed framework. With strong labels, our framework is able to achieve
state-of-the-art results in both datasets.
|
On-the-fly Approximation of Multivariate Total Variation Minimization | In the context of change-point detection, addressed by Total Variation
minimization strategies, an efficient on-the-fly algorithm has been designed
leading to exact solutions for univariate data. In this contribution, an
extension of such an on-the-fly strategy to multivariate data is investigated.
The proposed algorithm relies on the local validation of the Karush-Kuhn-Tucker
conditions on the dual problem. Showing that the non-local nature of the
multivariate setting precludes to obtain an exact on-the-fly solution, we
devise an on-the-fly algorithm delivering an approximate solution, whose
quality is controlled by a practitioner-tunable parameter, acting as a
trade-off between quality and computational cost. Performance assessment shows
that high quality solutions are obtained on-the-fly while benefiting of
computational costs several orders of magnitude lower than standard iterative
procedures. The proposed algorithm thus provides practitioners with an
efficient multivariate change-point detection on-the-fly procedure.
|
Spectral Norm of Random Kernel Matrices with Applications to Privacy | Kernel methods are an extremely popular set of techniques used for many
important machine learning and data analysis applications. In addition to
having good practical performances, these methods are supported by a
well-developed theory. Kernel methods use an implicit mapping of the input data
into a high dimensional feature space defined by a kernel function, i.e., a
function returning the inner product between the images of two data points in
the feature space. Central to any kernel method is the kernel matrix, which is
built by evaluating the kernel function on a given sample dataset.
In this paper, we initiate the study of non-asymptotic spectral theory of
random kernel matrices. These are n x n random matrices whose (i,j)th entry is
obtained by evaluating the kernel function on $x_i$ and $x_j$, where
$x_1,...,x_n$ are a set of n independent random high-dimensional vectors. Our
main contribution is to obtain tight upper bounds on the spectral norm (largest
eigenvalue) of random kernel matrices constructed by commonly used kernel
functions based on polynomials and Gaussian radial basis.
As an application of these results, we provide lower bounds on the distortion
needed for releasing the coefficients of kernel ridge regression under
attribute privacy, a general privacy notion which captures a large class of
privacy definitions. Kernel ridge regression is standard method for performing
non-parametric regression that regularly outperforms traditional regression
approaches in various domains. Our privacy distortion lower bounds are the
first for any kernel technique, and our analysis assumes realistic scenarios
for the input, unlike all previous lower bounds for other release problems
which only hold under very restrictive input settings.
|
svcR: An R Package for Support Vector Clustering improved with Geometric
Hashing applied to Lexical Pattern Discovery | We present a new R package which takes a numerical matrix format as data
input, and computes clusters using a support vector clustering method (SVC). We
have implemented an original 2D-grid labeling approach to speed up cluster
extraction. In this sense, SVC can be seen as an efficient cluster extraction
if clusters are separable in a 2-D map. Secondly we showed that this SVC
approach using a Jaccard-Radial base kernel can help to classify well enough a
set of terms into ontological classes and help to define regular expression
rules for information extraction in documents; our case study concerns a set of
terms and documents about developmental and molecular biology.
|
Collectively Embedding Multi-Relational Data for Predicting User
Preferences | Matrix factorization has found incredible success and widespread application
as a collaborative filtering based approach to recommendations. Unfortunately,
incorporating additional sources of evidence, especially ones that are
incomplete and noisy, is quite difficult to achieve in such models, however, is
often crucial for obtaining further gains in accuracy. For example, additional
information about businesses from reviews, categories, and attributes should be
leveraged for predicting user preferences, even though this information is
often inaccurate and partially-observed. Instead of creating customized methods
that are specific to each type of evidences, in this paper we present a generic
approach to factorization of relational data that collectively models all the
relations in the database. By learning a set of embeddings that are shared
across all the relations, the model is able to incorporate observed information
from all the relations, while also predicting all the relations of interest.
Our evaluation on multiple Amazon and Yelp datasets demonstrates effective
utilization of additional information for held-out preference prediction, but
further, we present accurate models even for the cold-starting businesses and
products for which we do not observe any ratings or reviews. We also illustrate
the capability of the model in imputing missing information and jointly
visualizing words, categories, and attribute factors.
|
A new approach for physiological time series | We developed a new approach for the analysis of physiological time series. An
iterative convolution filter is used to decompose the time series into various
components. Statistics of these components are extracted as features to
characterize the mechanisms underlying the time series. Motivated by the
studies that show many normal physiological systems involve irregularity while
the decrease of irregularity usually implies the abnormality, the statistics
for "outliers" in the components are used as features measuring irregularity.
Support vector machines are used to select the most relevant features that are
able to differentiate the time series from normal and abnormal systems. This
new approach is successfully used in the study of congestive heart failure by
heart beat interval time series.
|
Regularization-free estimation in trace regression with symmetric
positive semidefinite matrices | Over the past few years, trace regression models have received considerable
attention in the context of matrix completion, quantum state tomography, and
compressed sensing. Estimation of the underlying matrix from
regularization-based approaches promoting low-rankedness, notably nuclear norm
regularization, have enjoyed great popularity. In the present paper, we argue
that such regularization may no longer be necessary if the underlying matrix is
symmetric positive semidefinite (\textsf{spd}) and the design satisfies certain
conditions. In this situation, simple least squares estimation subject to an
\textsf{spd} constraint may perform as well as regularization-based approaches
with a proper choice of the regularization parameter, which entails knowledge
of the noise level and/or tuning. By contrast, constrained least squares
estimation comes without any tuning parameter and may hence be preferred due to
its simplicity.
|
Analysis of Stopping Active Learning based on Stabilizing Predictions | Within the natural language processing (NLP) community, active learning has
been widely investigated and applied in order to alleviate the annotation
bottleneck faced by developers of new NLP systems and technologies. This paper
presents the first theoretical analysis of stopping active learning based on
stabilizing predictions (SP). The analysis has revealed three elements that are
central to the success of the SP method: (1) bounds on Cohen's Kappa agreement
between successively trained models impose bounds on differences in F-measure
performance of the models; (2) since the stop set does not have to be labeled,
it can be made large in practice, helping to guarantee that the results
transfer to previously unseen streams of examples at test/application time; and
(3) good (low variance) sample estimates of Kappa between successive models can
be obtained. Proofs of relationships between the level of Kappa agreement and
the difference in performance between consecutive models are presented.
Specifically, if the Kappa agreement between two models exceeds a threshold T
(where $T>0$), then the difference in F-measure performance between those
models is bounded above by $\frac{4(1-T)}{T}$ in all cases. If precision of the
positive conjunction of the models is assumed to be $p$, then the bound can be
tightened to $\frac{4(1-T)}{(p+1)T}$.
|
Strategic Teaching and Learning in Games | It is known that there are uncoupled learning heuristics leading to Nash
equilibrium in all finite games. Why should players use such learning
heuristics and where could they come from? We show that there is no uncoupled
learning heuristic leading to Nash equilibrium in all finite games that a
player has an incentive to adopt, that would be evolutionary stable or that
could "learn itself". Rather, a player has an incentive to strategically teach
such a learning opponent in order secure at least the Stackelberg leader
payoff. The impossibility result remains intact when restricted to the classes
of generic games, two-player games, potential games, games with strategic
complements or 2x2 games, in which learning is known to be "nice". More
generally, it also applies to uncoupled learning heuristics leading to
correlated equilibria, rationalizable outcomes, iterated admissible outcomes,
or minimal curb sets. A possibility result restricted to "strategically
trivial" games fails if some generic games outside this class are considered as
well.
|
Use of Ensembles of Fourier Spectra in Capturing Recurrent Concepts in
Data Streams | In this research, we apply ensembles of Fourier encoded spectra to capture
and mine recurring concepts in a data stream environment. Previous research
showed that compact versions of Decision Trees can be obtained by applying the
Discrete Fourier Transform to accurately capture recurrent concepts in a data
stream. However, in highly volatile environments where new concepts emerge
often, the approach of encoding each concept in a separate spectrum is no
longer viable due to memory overload and thus in this research we present an
ensemble approach that addresses this problem. Our empirical results on real
world data and synthetic data exhibiting varying degrees of recurrence reveal
that the ensemble approach outperforms the single spectrum approach in terms of
classification accuracy, memory and execution time.
|
Social Trust Prediction via Max-norm Constrained 1-bit Matrix Completion | Social trust prediction addresses the significant problem of exploring
interactions among users in social networks. Naturally, this problem can be
formulated in the matrix completion framework, with each entry indicating the
trustness or distrustness. However, there are two challenges for the social
trust problem: 1) the observed data are with sign (1-bit) measurements; 2) they
are typically sampled non-uniformly. Most of the previous matrix completion
methods do not well handle the two issues. Motivated by the recent progress of
max-norm, we propose to solve the problem with a 1-bit max-norm constrained
formulation. Since max-norm is not easy to optimize, we utilize a reformulation
of max-norm which facilitates an efficient projected gradient decent algorithm.
We demonstrate the superiority of our formulation on two benchmark datasets.
|
Discriminative Switching Linear Dynamical Systems applied to
Physiological Condition Monitoring | We present a Discriminative Switching Linear Dynamical System (DSLDS) applied
to patient monitoring in Intensive Care Units (ICUs). Our approach is based on
identifying the state-of-health of a patient given their observed vital signs
using a discriminative classifier, and then inferring their underlying
physiological values conditioned on this status. The work builds on the
Factorial Switching Linear Dynamical System (FSLDS) (Quinn et al., 2009) which
has been previously used in a similar setting. The FSLDS is a generative model,
whereas the DSLDS is a discriminative model. We demonstrate on two real-world
datasets that the DSLDS is able to outperform the FSLDS in most cases of
interest, and that an $\alpha$-mixture of the two models achieves higher
performance than either of the two models separately.
|
Sampling Correctors | In many situations, sample data is obtained from a noisy or imperfect source.
In order to address such corruptions, this paper introduces the concept of a
sampling corrector. Such algorithms use structure that the distribution is
purported to have, in order to allow one to make "on-the-fly" corrections to
samples drawn from probability distributions. These algorithms then act as
filters between the noisy data and the end user.
We show connections between sampling correctors, distribution learning
algorithms, and distribution property testing algorithms. We show that these
connections can be utilized to expand the applicability of known distribution
learning and property testing algorithms as well as to achieve improved
algorithms for those tasks.
As a first step, we show how to design sampling correctors using proper
learning algorithms. We then focus on the question of whether algorithms for
sampling correctors can be more efficient in terms of sample complexity than
learning algorithms for the analogous families of distributions. When
correcting monotonicity, we show that this is indeed the case when also granted
query access to the cumulative distribution function. We also obtain sampling
correctors for monotonicity without this stronger type of access, provided that
the distribution be originally very close to monotone (namely, at a distance
$O(1/\log^2 n)$). In addition to that, we consider a restricted error model
that aims at capturing "missing data" corruptions. In this model, we show that
distributions that are close to monotone have sampling correctors that are
significantly more efficient than achievable by the learning approach.
We also consider the question of whether an additional source of independent
random bits is required by sampling correctors to implement the correction
process.
|
Classifying Relations by Ranking with Convolutional Neural Networks | Relation classification is an important semantic processing task for which
state-ofthe-art systems still rely on costly handcrafted features. In this work
we tackle the relation classification task using a convolutional neural network
that performs classification by ranking (CR-CNN). We propose a new pairwise
ranking loss function that makes it easy to reduce the impact of artificial
classes. We perform experiments using the the SemEval-2010 Task 8 dataset,
which is designed for the task of classifying the relationship between two
nominals marked in a sentence. Using CRCNN, we outperform the state-of-the-art
for this dataset and achieve a F1 of 84.1 without using any costly handcrafted
features. Additionally, our experimental results show that: (1) our approach is
more effective than CNN followed by a softmax classifier; (2) omitting the
representation of the artificial class Other improves both precision and
recall; and (3) using only word embeddings as input features is enough to
achieve state-of-the-art results if we consider only the text between the two
target nominals.
|
Handling oversampling in dynamic networks using link prediction | Oversampling is a common characteristic of data representing dynamic
networks. It introduces noise into representations of dynamic networks, but
there has been little work so far to compensate for it. Oversampling can affect
the quality of many important algorithmic problems on dynamic networks,
including link prediction. Link prediction seeks to predict edges that will be
added to the network given previous snapshots. We show that not only does
oversampling affect the quality of link prediction, but that we can use link
prediction to recover from the effects of oversampling. We also introduce a
novel generative model of noise in dynamic networks that represents
oversampling. We demonstrate the results of our approach on both synthetic and
real-world data.
|
Online Convex Optimization Using Predictions | Making use of predictions is a crucial, but under-explored, area of online
algorithms. This paper studies a class of online optimization problems where we
have external noisy predictions available. We propose a stochastic prediction
error model that generalizes prior models in the learning and stochastic
control communities, incorporates correlation among prediction errors, and
captures the fact that predictions improve as time passes. We prove that
achieving sublinear regret and constant competitive ratio for online algorithms
requires the use of an unbounded prediction window in adversarial settings, but
that under more realistic stochastic prediction error models it is possible to
use Averaging Fixed Horizon Control (AFHC) to simultaneously achieve sublinear
regret and constant competitive ratio in expectation using only a
constant-sized prediction window. Furthermore, we show that the performance of
AFHC is tightly concentrated around its mean.
|
Learning like a Child: Fast Novel Visual Concept Learning from Sentence
Descriptions of Images | In this paper, we address the task of learning novel visual concepts, and
their interactions with other concepts, from a few images with sentence
descriptions. Using linguistic context and visual features, our method is able
to efficiently hypothesize the semantic meaning of new words and add them to
its word dictionary so that they can be used to describe images which contain
these novel concepts. Our method has an image captioning module based on m-RNN
with several improvements. In particular, we propose a transposed weight
sharing scheme, which not only improves performance on image captioning, but
also makes the model more suitable for the novel concept learning task. We
propose methods to prevent overfitting the new concepts. In addition, three
novel concept datasets are constructed for this new task. In the experiments,
we show that our method effectively learns novel visual concepts from a few
examples without disturbing the previously learned concepts. The project page
is http://www.stat.ucla.edu/~junhua.mao/projects/child_learning.html
|
Complete Dictionary Recovery over the Sphere | We consider the problem of recovering a complete (i.e., square and
invertible) matrix $\mathbf A_0$, from $\mathbf Y \in \mathbb R^{n \times p}$
with $\mathbf Y = \mathbf A_0 \mathbf X_0$, provided $\mathbf X_0$ is
sufficiently sparse. This recovery problem is central to the theoretical
understanding of dictionary learning, which seeks a sparse representation for a
collection of input signals, and finds numerous applications in modern signal
processing and machine learning. We give the first efficient algorithm that
provably recovers $\mathbf A_0$ when $\mathbf X_0$ has $O(n)$ nonzeros per
column, under suitable probability model for $\mathbf X_0$. In contrast, prior
results based on efficient algorithms provide recovery guarantees when $\mathbf
X_0$ has only $O(n^{1-\delta})$ nonzeros per column for any constant $\delta
\in (0, 1)$.
Our algorithmic pipeline centers around solving a certain nonconvex
optimization problem with a spherical constraint, and hence is naturally
phrased in the language of manifold optimization. To show this apparently hard
problem is tractable, we first provide a geometric characterization of the
high-dimensional objective landscape, which shows that with high probability
there are no "spurious" local minima. This particular geometric structure
allows us to design a Riemannian trust region algorithm over the sphere that
provably converges to one local minimizer with an arbitrary initialization,
despite the presence of saddle points. The geometric approach we develop here
may also shed light on other problems arising from nonconvex recovery of
structured signals.
|
Max-margin Deep Generative Models | Deep generative models (DGMs) are effective on learning multilayered
representations of complex data and performing inference of input data by
exploring the generative ability. However, little work has been done on
examining or empowering the discriminative ability of DGMs on making accurate
predictions. This paper presents max-margin deep generative models (mmDGMs),
which explore the strongly discriminative principle of max-margin learning to
improve the discriminative power of DGMs, while retaining the generative
capability. We develop an efficient doubly stochastic subgradient algorithm for
the piecewise linear objective. Empirical results on MNIST and SVHN datasets
demonstrate that (1) max-margin learning can significantly improve the
prediction performance of DGMs and meanwhile retain the generative ability; and
(2) mmDGMs are competitive to the state-of-the-art fully discriminative
networks by employing deep convolutional neural networks (CNNs) as both
recognition and generative models.
|
Overlapping Communities Detection via Measure Space Embedding | We present a new algorithm for community detection. The algorithm uses random
walks to embed the graph in a space of measures, after which a modification of
$k$-means in that space is applied. The algorithm is therefore fast and easily
parallelizable. We evaluate the algorithm on standard random graph benchmarks,
including some overlapping community benchmarks, and find its performance to be
better or at least as good as previously known algorithms. We also prove a
linear time (in number of edges) guarantee for the algorithm on a
$p,q$-stochastic block model with $p \geq c\cdot N^{-\frac{1}{2} + \epsilon}$
and $p-q \geq c' \sqrt{p N^{-\frac{1}{2} + \epsilon} \log N}$.
|
Overlapping Community Detection by Online Cluster Aggregation | We present a new online algorithm for detecting overlapping communities. The
main ingredients are a modification of an online k-means algorithm and a new
approach to modelling overlap in communities. An evaluation on large benchmark
graphs shows that the quality of discovered communities compares favorably to
several methods in the recent literature, while the running time is
significantly improved.
|
Analysis of Nuclear Norm Regularization for Full-rank Matrix Completion | In this paper, we provide a theoretical analysis of the nuclear-norm
regularized least squares for full-rank matrix completion. Although similar
formulations have been examined by previous studies, their results are
unsatisfactory because only additive upper bounds are provided. Under the
assumption that the top eigenspaces of the target matrix are incoherent, we
derive a relative upper bound for recovering the best low-rank approximation of
the unknown matrix. Our relative upper bound is tighter than previous additive
bounds of other methods if the mass of the target matrix is concentrated on its
top eigenspaces, and also implies perfect recovery if it is low-rank. The
analysis is built upon the optimality condition of the regularized formulation
and existing guarantees for low-rank matrix completion. To the best of our
knowledge, this is first time such a relative bound is proved for the
regularized formulation of matrix completion.
|
Comparison of Training Methods for Deep Neural Networks | This report describes the difficulties of training neural networks and in
particular deep neural networks. It then provides a literature review of
training methods for deep neural networks, with a focus on pre-training. It
focuses on Deep Belief Networks composed of Restricted Boltzmann Machines and
Stacked Autoencoders and provides an outreach on further and alternative
approaches. It also includes related practical recommendations from the
literature on training them. In the second part, initial experiments using some
of the covered methods are performed on two databases. In particular,
experiments are performed on the MNIST hand-written digit dataset and on facial
emotion data from a Kaggle competition. The results are discussed in the
context of results reported in other research papers. An error rate lower than
the best contribution to the Kaggle competition is achieved using an optimized
Stacked Autoencoder.
|
Assessing binary classifiers using only positive and unlabeled data | Assessing the performance of a learned model is a crucial part of machine
learning. However, in some domains only positive and unlabeled examples are
available, which prohibits the use of most standard evaluation metrics. We
propose an approach to estimate any metric based on contingency tables,
including ROC and PR curves, using only positive and unlabeled data. Estimating
these performance metrics is essentially reduced to estimating the fraction of
(latent) positives in the unlabeled set, assuming known positives are a random
sample of all positives. We provide theoretical bounds on the quality of our
estimates, illustrate the importance of estimating the fraction of positives in
the unlabeled set and demonstrate empirically that we are able to reliably
estimate ROC and PR curves on real data.
|
FlowNet: Learning Optical Flow with Convolutional Networks | Convolutional neural networks (CNNs) have recently been very successful in a
variety of computer vision tasks, especially on those linked to recognition.
Optical flow estimation has not been among the tasks where CNNs were
successful. In this paper we construct appropriate CNNs which are capable of
solving the optical flow estimation problem as a supervised learning task. We
propose and compare two architectures: a generic architecture and another one
including a layer that correlates feature vectors at different image locations.
Since existing ground truth data sets are not sufficiently large to train a
CNN, we generate a synthetic Flying Chairs dataset. We show that networks
trained on this unrealistic data still generalize very well to existing
datasets such as Sintel and KITTI, achieving competitive accuracy at frame
rates of 5 to 10 fps.
|
Autonomy and Reliability of Continuous Active Learning for
Technology-Assisted Review | We enhance the autonomy of the continuous active learning method shown by
Cormack and Grossman (SIGIR 2014) to be effective for technology-assisted
review, in which documents from a collection are retrieved and reviewed, using
relevance feedback, until substantially all of the relevant documents have been
reviewed. Autonomy is enhanced through the elimination of topic-specific and
dataset-specific tuning parameters, so that the sole input required by the user
is, at the outset, a short query, topic description, or single relevant
document; and, throughout the review, ongoing relevance assessments of the
retrieved documents. We show that our enhancements consistently yield superior
results to Cormack and Grossman's version of continuous active learning, and
other methods, not only on average, but on the vast majority of topics from
four separate sets of tasks: the legal datasets examined by Cormack and
Grossman, the Reuters RCV1-v2 subject categories, the TREC 6 AdHoc task, and
the construction of the TREC 2002 filtering test collection.
|
Linear Spatial Pyramid Matching Using Non-convex and non-negative Sparse
Coding for Image Classification | Recently sparse coding have been highly successful in image classification
mainly due to its capability of incorporating the sparsity of image
representation. In this paper, we propose an improved sparse coding model based
on linear spatial pyramid matching(SPM) and Scale Invariant Feature Transform
(SIFT ) descriptors. The novelty is the simultaneous non-convex and
non-negative characters added to the sparse coding model. Our numerical
experiments show that the improved approach using non-convex and non-negative
sparse coding is superior than the original ScSPM[1] on several typical
databases.
|
Algorithms with Logarithmic or Sublinear Regret for Constrained
Contextual Bandits | We study contextual bandits with budget and time constraints, referred to as
constrained contextual bandits.The time and budget constraints significantly
complicate the exploration and exploitation tradeoff because they introduce
complex coupling among contexts over time.Such coupling effects make it
difficult to obtain oracle solutions that assume known statistics of bandits.
To gain insight, we first study unit-cost systems with known context
distribution. When the expected rewards are known, we develop an approximation
of the oracle, referred to Adaptive-Linear-Programming (ALP), which achieves
near-optimality and only requires the ordering of expected rewards. With these
highly desirable features, we then combine ALP with the upper-confidence-bound
(UCB) method in the general case where the expected rewards are unknown {\it a
priori}. We show that the proposed UCB-ALP algorithm achieves logarithmic
regret except for certain boundary cases. Further, we design algorithms and
obtain similar regret analysis results for more general systems with unknown
context distribution and heterogeneous costs. To the best of our knowledge,
this is the first work that shows how to achieve logarithmic regret in
constrained contextual bandits. Moreover, this work also sheds light on the
study of computationally efficient algorithms for general constrained
contextual bandits.
|
Random Forest for the Contextual Bandit Problem - extended version | To address the contextual bandit problem, we propose an online random forest
algorithm. The analysis of the proposed algorithm is based on the sample
complexity needed to find the optimal decision stump. Then, the decision stumps
are assembled in a random collection of decision trees, Bandit Forest. We show
that the proposed algorithm is optimal up to logarithmic factors. The
dependence of the sample complexity upon the number of contextual variables is
logarithmic. The computational cost of the proposed algorithm with respect to
the time horizon is linear. These analytical results allow the proposed
algorithm to be efficient in real applications, where the number of events to
process is huge, and where we expect that some contextual variables, chosen
from a large set, have potentially non- linear dependencies with the rewards.
In the experiments done to illustrate the theoretical analysis, Bandit Forest
obtain promising results in comparison with state-of-the-art algorithms.
|
Accelerated kernel discriminant analysis | In this paper, using a novel matrix factorization and simultaneous reduction
to diagonal form approach (or in short simultaneous reduction approach),
Accelerated Kernel Discriminant Analysis (AKDA) and Accelerated Kernel Subclass
Discriminant Analysis (AKSDA) are proposed. Specifically, instead of performing
the simultaneous reduction of the between- and within-class or subclass scatter
matrices, the nonzero eigenpairs (NZEP) of the so-called core matrix, which is
of relatively small dimensionality, and the Cholesky factorization of the
kernel matrix are computed, achieving more than one order of magnitude speed up
over kernel discriminant analysis (KDA). Moreover, consisting of a few
elementary matrix operations and very stable numerical algorithms, AKDA and
AKSDA offer improved classification accuracy. The experimental evaluation on
various datasets confirms that the proposed approaches provide state-of-the-art
performance in terms of both training time and classification accuracy.
|
An Active Learning Based Approach For Effective Video Annotation And
Retrieval | Conventional multimedia annotation/retrieval systems such as Normalized
Continuous Relevance Model (NormCRM) [16] require a fully labeled training data
for a good performance. Active Learning, by determining an order for labeling
the training data, allows for a good performance even before the training data
is fully annotated. In this work we propose an active learning algorithm, which
combines a novel measure of sample uncertainty with a novel clustering-based
approach for determining sample density and diversity and integrate it with
NormCRM. The clusters are also iteratively refined to ensure both feature and
label-level agreement among samples. We show that our approach outperforms
multiple baselines both on a recent, open character animation dataset and on
the popular TRECVID corpus at both the tasks of annotation and text-based
retrieval of videos.
|
Fast Sampling for Bayesian Max-Margin Models | Bayesian max-margin models have shown superiority in various practical
applications, such as text categorization, collaborative prediction, social
network link prediction and crowdsourcing, and they conjoin the flexibility of
Bayesian modeling and predictive strengths of max-margin learning. However,
Monte Carlo sampling for these models still remains challenging, especially for
applications that involve large-scale datasets. In this paper, we present the
stochastic subgradient Hamiltonian Monte Carlo (HMC) methods, which are easy to
implement and computationally efficient. We show the approximate detailed
balance property of subgradient HMC which reveals a natural and validated
generalization of the ordinary HMC. Furthermore, we investigate the variants
that use stochastic subsampling and thermostats for better scalability and
mixing. Using stochastic subgradient Markov Chain Monte Carlo (MCMC), we
efficiently solve the posterior inference task of various Bayesian max-margin
models and extensive experimental results demonstrate the effectiveness of our
approach.
|
Meta learning of bounds on the Bayes classifier error | Meta learning uses information from base learners (e.g. classifiers or
estimators) as well as information about the learning problem to improve upon
the performance of a single base learner. For example, the Bayes error rate of
a given feature space, if known, can be used to aid in choosing a classifier,
as well as in feature selection and model selection for the base classifiers
and the meta classifier. Recent work in the field of f-divergence functional
estimation has led to the development of simple and rapidly converging
estimators that can be used to estimate various bounds on the Bayes error. We
estimate multiple bounds on the Bayes error using an estimator that applies
meta learning to slowly converging plug-in estimators to obtain the parametric
convergence rate. We compare the estimated bounds empirically on simulated data
and then estimate the tighter bounds on features extracted from an image patch
analysis of sunspot continuum and magnetogram images.
|
Spectral MLE: Top-$K$ Rank Aggregation from Pairwise Comparisons | This paper explores the preference-based top-$K$ rank aggregation problem.
Suppose that a collection of items is repeatedly compared in pairs, and one
wishes to recover a consistent ordering that emphasizes the top-$K$ ranked
items, based on partially revealed preferences. We focus on the
Bradley-Terry-Luce (BTL) model that postulates a set of latent preference
scores underlying all items, where the odds of paired comparisons depend only
on the relative scores of the items involved.
We characterize the minimax limits on identifiability of top-$K$ ranked
items, in the presence of random and non-adaptive sampling. Our results
highlight a separation measure that quantifies the gap of preference scores
between the $K^{\text{th}}$ and $(K+1)^{\text{th}}$ ranked items. The minimum
sample complexity required for reliable top-$K$ ranking scales inversely with
the separation measure irrespective of other preference distribution metrics.
To approach this minimax limit, we propose a nearly linear-time ranking scheme,
called \emph{Spectral MLE}, that returns the indices of the top-$K$ items in
accordance to a careful score estimate. In a nutshell, Spectral MLE starts with
an initial score estimate with minimal squared loss (obtained via a spectral
method), and then successively refines each component with the assistance of
coordinate-wise MLEs. Encouragingly, Spectral MLE allows perfect top-$K$ item
identification under minimal sample complexity. The practical applicability of
Spectral MLE is further corroborated by numerical experiments.
|
Correlational Neural Networks | Common Representation Learning (CRL), wherein different descriptions (or
views) of the data are embedded in a common subspace, is receiving a lot of
attention recently. Two popular paradigms here are Canonical Correlation
Analysis (CCA) based approaches and Autoencoder (AE) based approaches. CCA
based approaches learn a joint representation by maximizing correlation of the
views when projected to the common subspace. AE based methods learn a common
representation by minimizing the error of reconstructing the two views. Each of
these approaches has its own advantages and disadvantages. For example, while
CCA based approaches outperform AE based approaches for the task of transfer
learning, they are not as scalable as the latter. In this work we propose an AE
based approach called Correlational Neural Network (CorrNet), that explicitly
maximizes correlation among the views when projected to the common subspace.
Through a series of experiments, we demonstrate that the proposed CorrNet is
better than the above mentioned approaches with respect to its ability to learn
correlated common representations. Further, we employ CorrNet for several cross
language tasks and show that the representations learned using CorrNet perform
better than the ones learned using other state of the art approaches.
|
Sign Stable Random Projections for Large-Scale Learning | We study the use of "sign $\alpha$-stable random projections" (where
$0<\alpha\leq 2$) for building basic data processing tools in the context of
large-scale machine learning applications (e.g., classification, regression,
clustering, and near-neighbor search). After the processing by sign stable
random projections, the inner products of the processed data approximate
various types of nonlinear kernels depending on the value of $\alpha$. Thus,
this approach provides an effective strategy for approximating nonlinear
learning algorithms essentially at the cost of linear learning. When $\alpha
=2$, it is known that the corresponding nonlinear kernel is the arc-cosine
kernel. When $\alpha=1$, the procedure approximates the arc-cos-$\chi^2$ kernel
(under certain condition). When $\alpha\rightarrow0+$, it corresponds to the
resemblance kernel.
From practitioners' perspective, the method of sign $\alpha$-stable random
projections is ready to be tested for large-scale learning applications, where
$\alpha$ can be simply viewed as a tuning parameter. What is missing in the
literature is an extensive empirical study to show the effectiveness of sign
stable random projections, especially for $\alpha\neq 2$ or 1. The paper
supplies such a study on a wide variety of classification datasets. In
particular, we compare shoulder-by-shoulder sign stable random projections with
the recently proposed "0-bit consistent weighted sampling (CWS)" (Li 2015).
|
Surrogate regret bounds for generalized classification performance
metrics | We consider optimization of generalized performance metrics for binary
classification by means of surrogate losses. We focus on a class of metrics,
which are linear-fractional functions of the false positive and false negative
rates (examples of which include $F_{\beta}$-measure, Jaccard similarity
coefficient, AM measure, and many others). Our analysis concerns the following
two-step procedure. First, a real-valued function $f$ is learned by minimizing
a surrogate loss for binary classification on the training sample. It is
assumed that the surrogate loss is a strongly proper composite loss function
(examples of which include logistic loss, squared-error loss, exponential loss,
etc.). Then, given $f$, a threshold $\widehat{\theta}$ is tuned on a separate
validation sample, by direct optimization of the target performance metric. We
show that the regret of the resulting classifier (obtained from thresholding
$f$ on $\widehat{\theta}$) measured with respect to the target metric is
upperbounded by the regret of $f$ measured with respect to the surrogate loss.
We also extend our results to cover multilabel classification and provide
regret bounds for micro- and macro-averaging measures. Our findings are further
analyzed in a computational study on both synthetic and real data sets.
|
Private Disclosure of Information in Health Tele-monitoring | We present a novel framework, called Private Disclosure of Information (PDI),
which is aimed to prevent an adversary from inferring certain sensitive
information about subjects using the data that they disclosed during
communication with an intended recipient. We show cases where it is possible to
achieve perfect privacy regardless of the adversary's auxiliary knowledge while
preserving full utility of the information to the intended recipient and
provide sufficient conditions for such cases. We also demonstrate the
applicability of PDI on a real-world data set that simulates a health
tele-monitoring scenario.
|
Lexical Translation Model Using a Deep Neural Network Architecture | In this paper we combine the advantages of a model using global source
sentence contexts, the Discriminative Word Lexicon, and neural networks. By
using deep neural networks instead of the linear maximum entropy model in the
Discriminative Word Lexicon models, we are able to leverage dependencies
between different source words due to the non-linearity. Furthermore, the
models for different target words can share parameters and therefore data
sparsity problems are effectively reduced.
By using this approach in a state-of-the-art translation system, we can
improve the performance by up to 0.5 BLEU points for three different language
pairs on the TED translation task.
|
Deep Neural Networks Regularization for Structured Output Prediction | A deep neural network model is a powerful framework for learning
representations. Usually, it is used to learn the relation $x \to y$ by
exploiting the regularities in the input $x$. In structured output prediction
problems, $y$ is multi-dimensional and structural relations often exist between
the dimensions. The motivation of this work is to learn the output dependencies
that may lie in the output data in order to improve the prediction accuracy.
Unfortunately, feedforward networks are unable to exploit the relations between
the outputs. In order to overcome this issue, we propose in this paper a
regularization scheme for training neural networks for these particular tasks
using a multi-task framework. Our scheme aims at incorporating the learning of
the output representation $y$ in the training process in an unsupervised
fashion while learning the supervised mapping function $x \to y$.
We evaluate our framework on a facial landmark detection problem which is a
typical structured output task. We show over two public challenging datasets
(LFPW and HELEN) that our regularization scheme improves the generalization of
deep neural networks and accelerates their training. The use of unlabeled data
and label-only data is also explored, showing an additional improvement of the
results. We provide an opensource implementation
(https://github.com/sbelharbi/structured-output-ae) of our framework.
|
Differentially Private Release and Learning of Threshold Functions | We prove new upper and lower bounds on the sample complexity of $(\epsilon,
\delta)$ differentially private algorithms for releasing approximate answers to
threshold functions. A threshold function $c_x$ over a totally ordered domain
$X$ evaluates to $c_x(y) = 1$ if $y \le x$, and evaluates to $0$ otherwise. We
give the first nontrivial lower bound for releasing thresholds with
$(\epsilon,\delta)$ differential privacy, showing that the task is impossible
over an infinite domain $X$, and moreover requires sample complexity $n \ge
\Omega(\log^*|X|)$, which grows with the size of the domain. Inspired by the
techniques used to prove this lower bound, we give an algorithm for releasing
thresholds with $n \le 2^{(1+ o(1))\log^*|X|}$ samples. This improves the
previous best upper bound of $8^{(1 + o(1))\log^*|X|}$ (Beimel et al., RANDOM
'13).
Our sample complexity upper and lower bounds also apply to the tasks of
learning distributions with respect to Kolmogorov distance and of properly PAC
learning thresholds with differential privacy. The lower bound gives the first
separation between the sample complexity of properly learning a concept class
with $(\epsilon,\delta)$ differential privacy and learning without privacy. For
properly learning thresholds in $\ell$ dimensions, this lower bound extends to
$n \ge \Omega(\ell \cdot \log^*|X|)$.
To obtain our results, we give reductions in both directions from releasing
and properly learning thresholds and the simpler interior point problem. Given
a database $D$ of elements from $X$, the interior point problem asks for an
element between the smallest and largest elements in $D$. We introduce new
recursive constructions for bounding the sample complexity of the interior
point problem, as well as further reductions and techniques for proving
impossibility results for other basic problems in differential privacy.
|
Becoming the Expert - Interactive Multi-Class Machine Teaching | Compared to machines, humans are extremely good at classifying images into
categories, especially when they possess prior knowledge of the categories at
hand. If this prior information is not available, supervision in the form of
teaching images is required. To learn categories more quickly, people should
see important and representative images first, followed by less important
images later - or not at all. However, image-importance is individual-specific,
i.e. a teaching image is important to a student if it changes their overall
ability to discriminate between classes. Further, students keep learning, so
while image-importance depends on their current knowledge, it also varies with
time.
In this work we propose an Interactive Machine Teaching algorithm that
enables a computer to teach challenging visual concepts to a human. Our
adaptive algorithm chooses, online, which labeled images from a teaching set
should be shown to the student as they learn. We show that a teaching strategy
that probabilistically models the student's ability and progress, based on
their correct and incorrect answers, produces better 'experts'. We present
results using real human participants across several varied and challenging
real-world datasets.
|
Or's of And's for Interpretable Classification, with Application to
Context-Aware Recommender Systems | We present a machine learning algorithm for building classifiers that are
comprised of a small number of disjunctions of conjunctions (or's of and's). An
example of a classifier of this form is as follows: If X satisfies (x1 = 'blue'
AND x3 = 'middle') OR (x1 = 'blue' AND x2 = '<15') OR (x1 = 'yellow'), then we
predict that Y=1, ELSE predict Y=0. An attribute-value pair is called a literal
and a conjunction of literals is called a pattern. Models of this form have the
advantage of being interpretable to human experts, since they produce a set of
conditions that concisely describe a specific class. We present two
probabilistic models for forming a pattern set, one with a Beta-Binomial prior,
and the other with Poisson priors. In both cases, there are prior parameters
that the user can set to encourage the model to have a desired size and shape,
to conform with a domain-specific definition of interpretability. We provide
two scalable MAP inference approaches: a pattern level search, which involves
association rule mining, and a literal level search. We show stronger priors
reduce computation. We apply the Bayesian Or's of And's (BOA) model to predict
user behavior with respect to in-vehicle context-aware personalized recommender
systems.
|
Nearly Optimal Deterministic Algorithm for Sparse Walsh-Hadamard
Transform | For every fixed constant $\alpha > 0$, we design an algorithm for computing
the $k$-sparse Walsh-Hadamard transform of an $N$-dimensional vector $x \in
\mathbb{R}^N$ in time $k^{1+\alpha} (\log N)^{O(1)}$. Specifically, the
algorithm is given query access to $x$ and computes a $k$-sparse $\tilde{x} \in
\mathbb{R}^N$ satisfying $\|\tilde{x} - \hat{x}\|_1 \leq c \|\hat{x} -
H_k(\hat{x})\|_1$, for an absolute constant $c > 0$, where $\hat{x}$ is the
transform of $x$ and $H_k(\hat{x})$ is its best $k$-sparse approximation. Our
algorithm is fully deterministic and only uses non-adaptive queries to $x$
(i.e., all queries are determined and performed in parallel when the algorithm
starts).
An important technical tool that we use is a construction of nearly optimal
and linear lossless condensers which is a careful instantiation of the GUV
condenser (Guruswami, Umans, Vadhan, JACM 2009). Moreover, we design a
deterministic and non-adaptive $\ell_1/\ell_1$ compressed sensing scheme based
on general lossless condensers that is equipped with a fast reconstruction
algorithm running in time $k^{1+\alpha} (\log N)^{O(1)}$ (for the GUV-based
condenser) and is of independent interest. Our scheme significantly simplifies
and improves an earlier expander-based construction due to Berinde, Gilbert,
Indyk, Karloff, Strauss (Allerton 2008).
Our methods use linear lossless condensers in a black box fashion; therefore,
any future improvement on explicit constructions of such condensers would
immediately translate to improved parameters in our framework (potentially
leading to $k (\log N)^{O(1)}$ reconstruction time with a reduced exponent in
the poly-logarithmic factor, and eliminating the extra parameter $\alpha$).
Finally, by allowing the algorithm to use randomness, while still using
non-adaptive queries, the running time of the algorithm can be improved to
$\tilde{O}(k \log^3 N)$.
|
Evaluation of Explore-Exploit Policies in Multi-result Ranking Systems | We analyze the problem of using Explore-Exploit techniques to improve
precision in multi-result ranking systems such as web search, query
autocompletion and news recommendation. Adopting an exploration policy directly
online, without understanding its impact on the production system, may have
unwanted consequences - the system may sustain large losses, create user
dissatisfaction, or collect exploration data which does not help improve
ranking quality. An offline framework is thus necessary to let us decide what
policy and how we should apply in a production environment to ensure positive
outcome. Here, we describe such an offline framework.
Using the framework, we study a popular exploration policy - Thompson
sampling. We show that there are different ways of implementing it in
multi-result ranking systems, each having different semantic interpretation and
leading to different results in terms of sustained click-through-rate (CTR)
loss and expected model improvement. In particular, we demonstrate that
Thompson sampling can act as an online learner optimizing CTR, which in some
cases can lead to an interesting outcome: lift in CTR during exploration. The
observation is important for production systems as it suggests that one can get
both valuable exploration data to improve ranking performance on the long run,
and at the same time increase CTR while exploration lasts.
|
Explaining the Success of AdaBoost and Random Forests as Interpolating
Classifiers | There is a large literature explaining why AdaBoost is a successful
classifier. The literature on AdaBoost focuses on classifier margins and
boosting's interpretation as the optimization of an exponential likelihood
function. These existing explanations, however, have been pointed out to be
incomplete. A random forest is another popular ensemble method for which there
is substantially less explanation in the literature. We introduce a novel
perspective on AdaBoost and random forests that proposes that the two
algorithms work for similar reasons. While both classifiers achieve similar
predictive accuracy, random forests cannot be conceived as a direct
optimization procedure. Rather, random forests is a self-averaging,
interpolating algorithm which creates what we denote as a "spikey-smooth"
classifier, and we view AdaBoost in the same light. We conjecture that both
AdaBoost and random forests succeed because of this mechanism. We provide a
number of examples and some theoretical justification to support this
explanation. In the process, we question the conventional wisdom that suggests
that boosting algorithms for classification require regularization or early
stopping and should be limited to low complexity classes of learners, such as
decision stumps. We conclude that boosting should be used like random forests:
with large decision trees and without direct regularization or early stopping.
|
Dual Averaging on Compactly-Supported Distributions And Application to
No-Regret Learning on a Continuum | We consider an online learning problem on a continuum. A decision maker is
given a compact feasible set $S$, and is faced with the following sequential
problem: at iteration~$t$, the decision maker chooses a distribution $x^{(t)}
\in \Delta(S)$, then a loss function $\ell^{(t)} : S \to \mathbb{R}_+$ is
revealed, and the decision maker incurs expected loss $\langle \ell^{(t)},
x^{(t)} \rangle = \mathbb{E}_{s \sim x^{(t)}} \ell^{(t)}(s)$. We view the
problem as an online convex optimization problem on the space $\Delta(S)$ of
Lebesgue-continnuous distributions on $S$. We prove a general regret bound for
the Dual Averaging method on $L^2(S)$, then prove that dual averaging with
$\omega$-potentials (a class of strongly convex regularizers) achieves
sublinear regret when $S$ is uniformly fat (a condition weaker than convexity).
|
Market forecasting using Hidden Markov Models | Working on the daily closing prices and logreturns, in this paper we deal
with the use of Hidden Markov Models (HMMs) to forecast the price of the
EUR/USD Futures. The aim of our work is to understand how the HMMs describe
different financial time series depending on their structure. Subsequently, we
analyse the forecasting methods exposed in the previous literature, putting on
evidence their pros and cons.
|
ASTROMLSKIT: A New Statistical Machine Learning Toolkit: A Platform for
Data Analytics in Astronomy | Astroinformatics is a new impact area in the world of astronomy, occasionally
called the final frontier, where several astrophysicists, statisticians and
computer scientists work together to tackle various data intensive astronomical
problems. Exponential growth in the data volume and increased complexity of the
data augments difficult questions to the existing challenges. Classical
problems in Astronomy are compounded by accumulation of astronomical volume of
complex data, rendering the task of classification and interpretation
incredibly laborious. The presence of noise in the data makes analysis and
interpretation even more arduous. Machine learning algorithms and data analytic
techniques provide the right platform for the challenges posed by these
problems. A diverse range of open problem like star-galaxy separation,
detection and classification of exoplanets, classification of supernovae is
discussed. The focus of the paper is the applicability and efficacy of various
machine learning algorithms like K Nearest Neighbor (KNN), random forest (RF),
decision tree (DT), Support Vector Machine (SVM), Na\"ive Bayes and Linear
Discriminant Analysis (LDA) in analysis and inference of the decision theoretic
problems in Astronomy. The machine learning algorithms, integrated into
ASTROMLSKIT, a toolkit developed in the course of the work, have been used to
analyze HabCat data and supernovae data. Accuracy has been found to be
appreciably good.
|
Learning Contextualized Music Semantics from Tags via a Siamese Network | Music information retrieval faces a challenge in modeling contextualized
musical concepts formulated by a set of co-occurring tags. In this paper, we
investigate the suitability of our recently proposed approach based on a
Siamese neural network in fighting off this challenge. By means of tag features
and probabilistic topic models, the network captures contextualized semantics
from tags via unsupervised learning. This leads to a distributed semantics
space and a potential solution to the out of vocabulary problem which has yet
to be sufficiently addressed. We explore the nature of the resultant
music-based semantics and address computational needs. We conduct experiments
on three public music tag collections -namely, CAL500, MagTag5K and Million
Song Dataset- and compare our approach to a number of state-of-the-art
semantics learning approaches. Comparative results suggest that this approach
outperforms previous approaches in terms of semantic priming and music tag
completion.
|
Who Spoke What? A Latent Variable Framework for the Joint Decoding of
Multiple Speakers and their Keywords | In this paper, we present a latent variable (LV) framework to identify all
the speakers and their keywords given a multi-speaker mixture signal. We
introduce two separate LVs to denote active speakers and the keywords uttered.
The dependency of a spoken keyword on the speaker is modeled through a
conditional probability mass function. The distribution of the mixture signal
is expressed in terms of the LV mass functions and speaker-specific-keyword
models. The proposed framework admits stochastic models, representing the
probability density function of the observation vectors given that a particular
speaker uttered a specific keyword, as speaker-specific-keyword models. The LV
mass functions are estimated in a Maximum Likelihood framework using the
Expectation Maximization (EM) algorithm. The active speakers and their keywords
are detected as modes of the joint distribution of the two LVs. In mixture
signals, containing two speakers uttering the keywords simultaneously, the
proposed framework achieves an accuracy of 82% for detecting both the speakers
and their respective keywords, using Student's-t mixture models as
speaker-specific-keyword models.
|
A Deep Learning Model for Structured Outputs with High-order Interaction | Many real-world applications are associated with structured data, where not
only input but also output has interplay. However, typical classification and
regression models often lack the ability of simultaneously exploring high-order
interaction within input and that within output. In this paper, we present a
deep learning model aiming to generate a powerful nonlinear functional mapping
from structured input to structured output. More specifically, we propose to
integrate high-order hidden units, guided discriminative pretraining, and
high-order auto-encoders for this purpose. We evaluate the model with three
datasets, and obtain state-of-the-art performances among competitive methods.
Our current work focuses on structured output regression, which is a less
explored area, although the model can be extended to handle structured label
classification.
|
Note on Equivalence Between Recurrent Neural Network Time Series Models
and Variational Bayesian Models | We observe that the standard log likelihood training objective for a
Recurrent Neural Network (RNN) model of time series data is equivalent to a
variational Bayesian training objective, given the proper choice of generative
and inference models. This perspective may motivate extensions to both RNNs and
variational Bayesian models. We propose one such extension, where multiple
particles are used for the hidden state of an RNN, allowing a natural
representation of uncertainty or multimodality.
|
Semi-Orthogonal Multilinear PCA with Relaxed Start | Principal component analysis (PCA) is an unsupervised method for learning
low-dimensional features with orthogonal projections. Multilinear PCA methods
extend PCA to deal with multidimensional data (tensors) directly via
tensor-to-tensor projection or tensor-to-vector projection (TVP). However,
under the TVP setting, it is difficult to develop an effective multilinear PCA
method with the orthogonality constraint. This paper tackles this problem by
proposing a novel Semi-Orthogonal Multilinear PCA (SO-MPCA) approach. SO-MPCA
learns low-dimensional features directly from tensors via TVP by imposing the
orthogonality constraint in only one mode. This formulation results in more
captured variance and more learned features than full orthogonality. For better
generalization, we further introduce a relaxed start (RS) strategy to get
SO-MPCA-RS by fixing the starting projection vectors, which increases the bias
and reduces the variance of the learning model. Experiments on both face (2D)
and gait (3D) data demonstrate that SO-MPCA-RS outperforms other competing
algorithms on the whole, and the relaxed start strategy is also effective for
other TVP-based PCA methods.
|
Multi-user lax communications: a multi-armed bandit approach | Inspired by cognitive radio networks, we consider a setting where multiple
users share several channels modeled as a multi-user multi-armed bandit (MAB)
problem. The characteristics of each channel are unknown and are different for
each user. Each user can choose between the channels, but her success depends
on the particular channel chosen as well as on the selections of other users:
if two users select the same channel their messages collide and none of them
manages to send any data. Our setting is fully distributed, so there is no
central control. As in many communication systems, the users cannot set up a
direct communication protocol, so information exchange must be limited to a
minimum. We develop an algorithm for learning a stable configuration for the
multi-user MAB problem. We further offer both convergence guarantees and
experiments inspired by real communication networks, including comparison to
state-of-the-art algorithms.
|
Model Selection and Overfitting in Genetic Programming: Empirical Study
[Extended Version] | Genetic Programming has been very successful in solving a large area of
problems but its use as a machine learning algorithm has been limited so far.
One of the reasons is the problem of overfitting which cannot be solved or
suppresed as easily as in more traditional approaches. Another problem, closely
related to overfitting, is the selection of the final model from the
population.
In this article we present our research that addresses both problems:
overfitting and model selection. We compare several ways of dealing with
ovefitting, based on Random Sampling Technique (RST) and on using a validation
set, all with an emphasis on model selection. We subject each approach to a
thorough testing on artificial and real--world datasets and compare them with
the standard approach, which uses the full training data, as a baseline.
|
Lateral Connections in Denoising Autoencoders Support Supervised
Learning | We show how a deep denoising autoencoder with lateral connections can be used
as an auxiliary unsupervised learning task to support supervised learning. The
proposed model is trained to minimize simultaneously the sum of supervised and
unsupervised cost functions by back-propagation, avoiding the need for
layer-wise pretraining. It improves the state of the art significantly in the
permutation-invariant MNIST classification task.
|
Hierarchical Subquery Evaluation for Active Learning on a Graph | To train good supervised and semi-supervised object classifiers, it is
critical that we not waste the time of the human experts who are providing the
training labels. Existing active learning strategies can have uneven
performance, being efficient on some datasets but wasteful on others, or
inconsistent just between runs on the same dataset. We propose perplexity based
graph construction and a new hierarchical subquery evaluation algorithm to
combat this variability, and to release the potential of Expected Error
Reduction.
Under some specific circumstances, Expected Error Reduction has been one of
the strongest-performing informativeness criteria for active learning. Until
now, it has also been prohibitively costly to compute for sizeable datasets. We
demonstrate our highly practical algorithm, comparing it to other active
learning measures on classification datasets that vary in sparsity,
dimensionality, and size. Our algorithm is consistent over multiple runs and
achieves high accuracy, while querying the human expert for labels at a
frequency that matches their desired time budget.
|
Deep Neural Networks with Random Gaussian Weights: A Universal
Classification Strategy? | Three important properties of a classification machinery are: (i) the system
preserves the core information of the input data; (ii) the training examples
convey information about unseen data; and (iii) the system is able to treat
differently points from different classes. In this work we show that these
fundamental properties are satisfied by the architecture of deep neural
networks. We formally prove that these networks with random Gaussian weights
perform a distance-preserving embedding of the data, with a special treatment
for in-class and out-of-class data. Similar points at the input of the network
are likely to have a similar output. The theoretical analysis of deep networks
here presented exploits tools used in the compressed sensing and dictionary
learning literature, thereby making a formal connection between these important
topics. The derived results allow drawing conclusions on the metric learning
properties of the network and their relation to its structure, as well as
providing bounds on the required size of the training set such that the
training examples would represent faithfully the unseen data. The results are
validated with state-of-the-art trained networks.
|
On the Structure, Covering, and Learning of Poisson Multinomial
Distributions | An $(n,k)$-Poisson Multinomial Distribution (PMD) is the distribution of the
sum of $n$ independent random vectors supported on the set ${\cal
B}_k=\{e_1,\ldots,e_k\}$ of standard basis vectors in $\mathbb{R}^k$. We prove
a structural characterization of these distributions, showing that, for all
$\varepsilon >0$, any $(n, k)$-Poisson multinomial random vector is
$\varepsilon$-close, in total variation distance, to the sum of a discretized
multidimensional Gaussian and an independent $(\text{poly}(k/\varepsilon),
k)$-Poisson multinomial random vector. Our structural characterization extends
the multi-dimensional CLT of Valiant and Valiant, by simultaneously applying to
all approximation requirements $\varepsilon$. In particular, it overcomes
factors depending on $\log n$ and, importantly, the minimum eigenvalue of the
PMD's covariance matrix from the distance to a multidimensional Gaussian random
variable.
We use our structural characterization to obtain an $\varepsilon$-cover, in
total variation distance, of the set of all $(n, k)$-PMDs, significantly
improving the cover size of Daskalakis and Papadimitriou, and obtaining the
same qualitative dependence of the cover size on $n$ and $\varepsilon$ as the
$k=2$ cover of Daskalakis and Papadimitriou. We further exploit this structure
to show that $(n,k)$-PMDs can be learned to within $\varepsilon$ in total
variation distance from $\tilde{O}_k(1/\varepsilon^2)$ samples, which is
near-optimal in terms of dependence on $\varepsilon$ and independent of $n$. In
particular, our result generalizes the single-dimensional result of Daskalakis,
Diakonikolas, and Servedio for Poisson Binomials to arbitrary dimension.
|
Thompson Sampling for Budgeted Multi-armed Bandits | Thompson sampling is one of the earliest randomized algorithms for
multi-armed bandits (MAB). In this paper, we extend the Thompson sampling to
Budgeted MAB, where there is random cost for pulling an arm and the total cost
is constrained by a budget. We start with the case of Bernoulli bandits, in
which the random rewards (costs) of an arm are independently sampled from a
Bernoulli distribution. To implement the Thompson sampling algorithm in this
case, at each round, we sample two numbers from the posterior distributions of
the reward and cost for each arm, obtain their ratio, select the arm with the
maximum ratio, and then update the posterior distributions. We prove that the
distribution-dependent regret bound of this algorithm is $O(\ln B)$, where $B$
denotes the budget. By introducing a Bernoulli trial, we further extend this
algorithm to the setting that the rewards (costs) are drawn from general
distributions, and prove that its regret bound remains almost the same. Our
simulation results demonstrate the effectiveness of the proposed algorithm.
|
Theory of Optimizing Pseudolinear Performance Measures: Application to
F-measure | Non-linear performance measures are widely used for the evaluation of
learning algorithms. For example, $F$-measure is a commonly used performance
measure for classification problems in machine learning and information
retrieval community. We study the theoretical properties of a subset of
non-linear performance measures called pseudo-linear performance measures which
includes $F$-measure, \emph{Jaccard Index}, among many others. We establish
that many notions of $F$-measures and \emph{Jaccard Index} are pseudo-linear
functions of the per-class false negatives and false positives for binary,
multiclass and multilabel classification. Based on this observation, we present
a general reduction of such performance measure optimization problem to
cost-sensitive classification problem with unknown costs. We then propose an
algorithm with provable guarantees to obtain an approximately optimal
classifier for the $F$-measure by solving a series of cost-sensitive
classification problems. The strength of our analysis is to be valid on any
dataset and any class of classifiers, extending the existing theoretical
results on pseudo-linear measures, which are asymptotic in nature. We also
establish the multi-objective nature of the $F$-score maximization problem by
linking the algorithm with the weighted-sum approach used in multi-objective
optimization. We present numerical experiments to illustrate the relative
importance of cost asymmetry and thresholding when learning linear classifiers
on various $F$-measure optimization tasks.
|
Grounded Discovery of Coordinate Term Relationships between Software
Entities | We present an approach for the detection of coordinate-term relationships
between entities from the software domain, that refer to Java classes. Usually,
relations are found by examining corpus statistics associated with text
entities. In some technical domains, however, we have access to additional
information about the real-world objects named by the entities, suggesting that
coupling information about the "grounded" entities with corpus statistics might
lead to improved methods for relation discovery. To this end, we develop a
similarity measure for Java classes using distributional information about how
they are used in software, which we combine with corpus statistics on the
distribution of contexts in which the classes appear in text. Using our
approach, cross-validation accuracy on this dataset can be improved
dramatically, from around 60% to 88%. Human labeling results show that our
classifier has an F1 score of 86% over the top 1000 predicted pairs.
|
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.