title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Stochastic Neural Networks with Monotonic Activation Functions | We propose a Laplace approximation that creates a stochastic unit from any
smooth monotonic activation function, using only Gaussian noise. This paper
investigates the application of this stochastic approximation in training a
family of Restricted Boltzmann Machines (RBM) that are closely linked to
Bregman divergences. This family, that we call exponential family RBM
(Exp-RBM), is a subset of the exponential family Harmoniums that expresses
family members through a choice of smooth monotonic non-linearity for each
neuron. Using contrastive divergence along with our Gaussian approximation, we
show that Exp-RBM can learn useful representations using novel stochastic
units.
|
Practical Algorithms for Learning Near-Isometric Linear Embeddings | We propose two practical non-convex approaches for learning near-isometric,
linear embeddings of finite sets of data points. Given a set of training points
$\mathcal{X}$, we consider the secant set $S(\mathcal{X})$ that consists of all
pairwise difference vectors of $\mathcal{X}$, normalized to lie on the unit
sphere. The problem can be formulated as finding a symmetric and positive
semi-definite matrix $\boldsymbol{\Psi}$ that preserves the norms of all the
vectors in $S(\mathcal{X})$ up to a distortion parameter $\delta$. Motivated by
non-negative matrix factorization, we reformulate our problem into a Frobenius
norm minimization problem, which is solved by the Alternating Direction Method
of Multipliers (ADMM) and develop an algorithm, FroMax. Another method solves
for a projection matrix $\boldsymbol{\Psi}$ by minimizing the restricted
isometry property (RIP) directly over the set of symmetric, postive
semi-definite matrices. Applying ADMM and a Moreau decomposition on a proximal
mapping, we develop another algorithm, NILE-Pro, for dimensionality reduction.
FroMax is shown to converge faster for smaller $\delta$ while NILE-Pro
converges faster for larger $\delta$. Both non-convex approaches are then
empirically demonstrated to be more computationally efficient than prior convex
approaches for a number of applications in machine learning and signal
processing.
|
Supervised Dimensionality Reduction via Distance Correlation
Maximization | In our work, we propose a novel formulation for supervised dimensionality
reduction based on a nonlinear dependency criterion called Statistical Distance
Correlation, Szekely et. al. (2007). We propose an objective which is free of
distributional assumptions on regression variables and regression model
assumptions. Our proposed formulation is based on learning a low-dimensional
feature representation $\mathbf{z}$, which maximizes the squared sum of
Distance Correlations between low dimensional features $\mathbf{z}$ and
response $y$, and also between features $\mathbf{z}$ and covariates
$\mathbf{x}$. We propose a novel algorithm to optimize our proposed objective
using the Generalized Minimization Maximizaiton method of \Parizi et. al.
(2015). We show superior empirical results on multiple datasets proving the
effectiveness of our proposed approach over several relevant state-of-the-art
supervised dimensionality reduction methods.
|
Dimensionality-Dependent Generalization Bounds for $k$-Dimensional
Coding Schemes | The $k$-dimensional coding schemes refer to a collection of methods that
attempt to represent data using a set of representative $k$-dimensional
vectors, and include non-negative matrix factorization, dictionary learning,
sparse coding, $k$-means clustering and vector quantization as special cases.
Previous generalization bounds for the reconstruction error of the
$k$-dimensional coding schemes are mainly dimensionality independent. A major
advantage of these bounds is that they can be used to analyze the
generalization error when data is mapped into an infinite- or high-dimensional
feature space. However, many applications use finite-dimensional data features.
Can we obtain dimensionality-dependent generalization bounds for
$k$-dimensional coding schemes that are tighter than dimensionality-independent
bounds when data is in a finite-dimensional feature space? The answer is
positive. In this paper, we address this problem and derive a
dimensionality-dependent generalization bound for $k$-dimensional coding
schemes by bounding the covering number of the loss function class induced by
the reconstruction error. The bound is of order
$\mathcal{O}\left(\left(mk\ln(mkn)/n\right)^{\lambda_n}\right)$, where $m$ is
the dimension of features, $k$ is the number of the columns in the linear
implementation of coding schemes, $n$ is the size of sample, $\lambda_n>0.5$
when $n$ is finite and $\lambda_n=0.5$ when $n$ is infinite. We show that our
bound can be tighter than previous results, because it avoids inducing the
worst-case upper bound on $k$ of the loss function and converges faster. The
proposed generalization bound is also applied to some specific coding schemes
to demonstrate that the dimensionality-dependent bound is an indispensable
complement to these dimensionality-independent generalization bounds.
|
A Unified Approach for Learning the Parameters of Sum-Product Networks | We present a unified approach for learning the parameters of Sum-Product
networks (SPNs). We prove that any complete and decomposable SPN is equivalent
to a mixture of trees where each tree corresponds to a product of univariate
distributions. Based on the mixture model perspective, we characterize the
objective function when learning SPNs based on the maximum likelihood
estimation (MLE) principle and show that the optimization problem can be
formulated as a signomial program. We construct two parameter learning
algorithms for SPNs by using sequential monomial approximations (SMA) and the
concave-convex procedure (CCCP), respectively. The two proposed methods
naturally admit multiplicative updates, hence effectively avoiding the
projection operation. With the help of the unified framework, we also show
that, in the case of SPNs, CCCP leads to the same algorithm as Expectation
Maximization (EM) despite the fact that they are different in general.
|
Sparse Diffusion Steepest-Descent for One Bit Compressed Sensing in
Wireless Sensor Networks | This letter proposes a sparse diffusion steepest-descent algorithm for one
bit compressed sensing in wireless sensor networks. The approach exploits the
diffusion strategy from distributed learning in the one bit compressed sensing
framework. To estimate a common sparse vector cooperatively from only the sign
of measurements, steepest-descent is used to minimize the suitable global and
local convex cost functions. A diffusion strategy is suggested for distributive
learning of the sparse vector. Simulation results show the effectiveness of the
proposed distributed algorithm compared to the state-of-the-art non
distributive algorithms in the one bit compressed sensing framework.
|
On the Reducibility of Submodular Functions | The scalability of submodular optimization methods is critical for their
usability in practice. In this paper, we study the reducibility of submodular
functions, a property that enables us to reduce the solution space of
submodular optimization problems without performance loss. We introduce the
concept of reducibility using marginal gains. Then we show that by adding
perturbation, we can endow irreducible functions with reducibility, based on
which we propose the perturbation-reduction optimization framework. Our
theoretical analysis proves that given the perturbation scales, the
reducibility gain could be computed, and the performance loss has additive
upper bounds. We further conduct empirical studies and the results demonstrate
that our proposed framework significantly accelerates existing optimization
methods for irreducible submodular functions with a cost of only small
performance losses.
|
Fitting Spectral Decay with the $k$-Support Norm | The spectral $k$-support norm enjoys good estimation properties in low rank
matrix learning problems, empirically outperforming the trace norm. Its unit
ball is the convex hull of rank $k$ matrices with unit Frobenius norm. In this
paper we generalize the norm to the spectral $(k,p)$-support norm, whose
additional parameter $p$ can be used to tailor the norm to the decay of the
spectrum of the underlying model. We characterize the unit ball and we
explicitly compute the norm. We further provide a conditional gradient method
to solve regularization problems with the norm, and we derive an efficient
algorithm to compute the Euclidean projection on the unit ball in the case
$p=\infty$. In numerical experiments, we show that allowing $p$ to vary
significantly improves performance over the spectral $k$-support norm on
various matrix completion benchmarks, and better captures the spectral decay of
the underlying model.
|
Approximate Message Passing with Nearest Neighbor Sparsity Pattern
Learning | We consider the problem of recovering clustered sparse signals with no prior
knowledge of the sparsity pattern. Beyond simple sparsity, signals of interest
often exhibits an underlying sparsity pattern which, if leveraged, can improve
the reconstruction performance. However, the sparsity pattern is usually
unknown a priori. Inspired by the idea of k-nearest neighbor (k-NN) algorithm,
we propose an efficient algorithm termed approximate message passing with
nearest neighbor sparsity pattern learning (AMP-NNSPL), which learns the
sparsity pattern adaptively. AMP-NNSPL specifies a flexible spike and slab
prior on the unknown signal and, after each AMP iteration, sets the sparse
ratios as the average of the nearest neighbor estimates via expectation
maximization (EM). Experimental results on both synthetic and real data
demonstrate the superiority of our proposed algorithm both in terms of
reconstruction performance and computational complexity.
|
NFL Play Prediction | Based on NFL game data we try to predict the outcome of a play in multiple
different ways. An application of this is the following: by plugging in various
play options one could determine the best play for a given situation in real
time. While the outcome of a play can be described in many ways we had the most
promising results with a newly defined measure that we call "progress". We see
this work as a first step to include predictive analysis into NFL playcalling.
|
Robust Non-linear Regression: A Greedy Approach Employing Kernels with
Application to Image Denoising | We consider the task of robust non-linear regression in the presence of both
inlier noise and outliers. Assuming that the unknown non-linear function
belongs to a Reproducing Kernel Hilbert Space (RKHS), our goal is to estimate
the set of the associated unknown parameters. Due to the presence of outliers,
common techniques such as the Kernel Ridge Regression (KRR) or the Support
Vector Regression (SVR) turn out to be inadequate. Instead, we employ sparse
modeling arguments to explicitly model and estimate the outliers, adopting a
greedy approach. The proposed robust scheme, i.e., Kernel Greedy Algorithm for
Robust Denoising (KGARD), is inspired by the classical Orthogonal Matching
Pursuit (OMP) algorithm. Specifically, the proposed method alternates between a
KRR task and an OMP-like selection step. Theoretical results concerning the
identification of the outliers are provided. Moreover, KGARD is compared
against other cutting edge methods, where its performance is evaluated via a
set of experiments with various types of noise. Finally, the proposed robust
estimation framework is applied to the task of image denoising, and its
enhanced performance in the presence of outliers is demonstrated.
|
Scalable Models for Computing Hierarchies in Information Networks | Information hierarchies are organizational structures that often used to
organize and present large and complex information as well as provide a
mechanism for effective human navigation. Fortunately, many statistical and
computational models exist that automatically generate hierarchies; however,
the existing approaches do not consider linkages in information {\em networks}
that are increasingly common in real-world scenarios. Current approaches also
tend to present topics as an abstract probably distribution over words, etc
rather than as tangible nodes from the original network. Furthermore, the
statistical techniques present in many previous works are not yet capable of
processing data at Web-scale. In this paper we present the Hierarchical
Document Topic Model (HDTM), which uses a distributed vertex-programming
process to calculate a nonparametric Bayesian generative model. Experiments on
three medium size data sets and the entire Wikipedia dataset show that HDTM can
infer accurate hierarchies even over large information networks.
|
Variational Inference: A Review for Statisticians | One of the core problems of modern statistics is to approximate
difficult-to-compute probability densities. This problem is especially
important in Bayesian statistics, which frames all inference about unknown
quantities as a calculation involving the posterior density. In this paper, we
review variational inference (VI), a method from machine learning that
approximates probability densities through optimization. VI has been used in
many applications and tends to be faster than classical methods, such as Markov
chain Monte Carlo sampling. The idea behind VI is to first posit a family of
densities and then to find the member of that family which is close to the
target. Closeness is measured by Kullback-Leibler divergence. We review the
ideas behind mean-field variational inference, discuss the special case of VI
applied to exponential family models, present a full example with a Bayesian
mixture of Gaussians, and derive a variant that uses stochastic optimization to
scale up to massive data. We discuss modern research in VI and highlight
important open problems. VI is powerful, but it is not yet well understood. Our
hope in writing this paper is to catalyze statistical research on this class of
algorithms.
|
Nonlinear Hebbian learning as a unifying principle in receptive field
formation | The development of sensory receptive fields has been modeled in the past by a
variety of models including normative models such as sparse coding or
independent component analysis and bottom-up models such as spike-timing
dependent plasticity or the Bienenstock-Cooper-Munro model of synaptic
plasticity. Here we show that the above variety of approaches can all be
unified into a single common principle, namely Nonlinear Hebbian Learning. When
Nonlinear Hebbian Learning is applied to natural images, receptive field shapes
were strongly constrained by the input statistics and preprocessing, but
exhibited only modest variation across different choices of nonlinearities in
neuron models or synaptic plasticity rules. Neither overcompleteness nor sparse
network activity are necessary for the development of localized receptive
fields. The analysis of alternative sensory modalities such as auditory models
or V2 development lead to the same conclusions. In all examples, receptive
fields can be predicted a priori by reformulating an abstract model as
nonlinear Hebbian learning. Thus nonlinear Hebbian learning and natural
statistics can account for many aspects of receptive field formation across
models and sensory modalities.
|
Weakly-supervised Disentangling with Recurrent Transformations for 3D
View Synthesis | An important problem for both graphics and vision is to synthesize novel
views of a 3D object from a single image. This is particularly challenging due
to the partial observability inherent in projecting a 3D object onto the image
space, and the ill-posedness of inferring object shape and pose. However, we
can train a neural network to address the problem if we restrict our attention
to specific object categories (in our case faces and chairs) for which we can
gather ample training data. In this paper, we propose a novel recurrent
convolutional encoder-decoder network that is trained end-to-end on the task of
rendering rotated objects starting from a single image. The recurrent structure
allows our model to capture long-term dependencies along a sequence of
transformations. We demonstrate the quality of its predictions for human faces
on the Multi-PIE dataset and for a dataset of 3D chair models, and also show
its ability to disentangle latent factors of variation (e.g., identity and
pose) without using full supervision.
|
Low-Rank Representation over the Manifold of Curves | In machine learning it is common to interpret each data point as a vector in
Euclidean space. However the data may actually be functional i.e.\ each data
point is a function of some variable such as time and the function is
discretely sampled. The naive treatment of functional data as traditional
multivariate data can lead to poor performance since the algorithms are
ignoring the correlation in the curvature of each function. In this paper we
propose a method to analyse subspace structure of the functional data by using
the state of the art Low-Rank Representation (LRR). Experimental evaluation on
synthetic and real data reveals that this method massively outperforms
conventional LRR in tasks concerning functional data.
|
Brain4Cars: Car That Knows Before You Do via Sensory-Fusion Deep
Learning Architecture | Advanced Driver Assistance Systems (ADAS) have made driving safer over the
last decade. They prepare vehicles for unsafe road conditions and alert drivers
if they perform a dangerous maneuver. However, many accidents are unavoidable
because by the time drivers are alerted, it is already too late. Anticipating
maneuvers beforehand can alert drivers before they perform the maneuver and
also give ADAS more time to avoid or prepare for the danger.
In this work we propose a vehicular sensor-rich platform and learning
algorithms for maneuver anticipation. For this purpose we equip a car with
cameras, Global Positioning System (GPS), and a computing device to capture the
driving context from both inside and outside of the car. In order to anticipate
maneuvers, we propose a sensory-fusion deep learning architecture which jointly
learns to anticipate and fuse multiple sensory streams. Our architecture
consists of Recurrent Neural Networks (RNNs) that use Long Short-Term Memory
(LSTM) units to capture long temporal dependencies. We propose a novel training
procedure which allows the network to predict the future given only a partial
temporal context. We introduce a diverse data set with 1180 miles of natural
freeway and city driving, and show that we can anticipate maneuvers 3.5 seconds
before they occur in real-time with a precision and recall of 90.5\% and 87.4\%
respectively.
|
Learning Preferences for Manipulation Tasks from Online Coactive
Feedback | We consider the problem of learning preferences over trajectories for mobile
manipulators such as personal robots and assembly line robots. The preferences
we learn are more intricate than simple geometric constraints on trajectories;
they are rather governed by the surrounding context of various objects and
human interactions in the environment. We propose a coactive online learning
framework for teaching preferences in contextually rich environments. The key
novelty of our approach lies in the type of feedback expected from the user:
the human user does not need to demonstrate optimal trajectories as training
data, but merely needs to iteratively provide trajectories that slightly
improve over the trajectory currently proposed by the system. We argue that
this coactive preference feedback can be more easily elicited than
demonstrations of optimal trajectories. Nevertheless, theoretical regret bounds
of our algorithm match the asymptotic rates of optimal trajectory algorithms.
We implement our algorithm on two high degree-of-freedom robots, PR2 and
Baxter, and present three intuitive mechanisms for providing such incremental
feedback. In our experimental evaluation we consider two context rich settings
-- household chores and grocery store checkout -- and show that users are able
to train the robot with just a few feedbacks (taking only a few
minutes).\footnote{Parts of this work has been published at NIPS and ISRR
conferences~\citep{Jain13,Jain13b}. This journal submission presents a
consistent full paper, and also includes the proof of regret bounds, more
details of the robotic system, and a thorough related work.}
|
End-to-End Relation Extraction using LSTMs on Sequences and Tree
Structures | We present a novel end-to-end neural model to extract entities and relations
between them. Our recurrent neural network based model captures both word
sequence and dependency tree substructure information by stacking bidirectional
tree-structured LSTM-RNNs on bidirectional sequential LSTM-RNNs. This allows
our model to jointly represent both entities and relations with shared
parameters in a single model. We further encourage detection of entities during
training and use of entity information in relation extraction via entity
pretraining and scheduled sampling. Our model improves over the
state-of-the-art feature-based model on end-to-end relation extraction,
achieving 12.1% and 5.7% relative error reductions in F1-score on ACE2005 and
ACE2004, respectively. We also show that our LSTM-RNN based model compares
favorably to the state-of-the-art CNN based model (in F1-score) on nominal
relation classification (SemEval-2010 Task 8). Finally, we present an extensive
ablation analysis of several model components.
|
Open challenges in understanding development and evolution of speech
forms: The roles of embodied self-organization, motivation and active
exploration | This article discusses open scientific challenges for understanding
development and evolution of speech forms, as a commentary to Moulin-Frier et
al. (Moulin-Frier et al., 2015). Based on the analysis of mathematical models
of the origins of speech forms, with a focus on their assumptions , we study
the fundamental question of how speech can be formed out of non--speech, at
both developmental and evolutionary scales. In particular, we emphasize the
importance of embodied self-organization , as well as the role of mechanisms of
motivation and active curiosity-driven exploration in speech formation. Finally
, we discuss an evolutionary-developmental perspective of the origins of
speech.
|
DrMAD: Distilling Reverse-Mode Automatic Differentiation for Optimizing
Hyperparameters of Deep Neural Networks | The performance of deep neural networks is well-known to be sensitive to the
setting of their hyperparameters. Recent advances in reverse-mode automatic
differentiation allow for optimizing hyperparameters with gradients. The
standard way of computing these gradients involves a forward and backward pass
of computations. However, the backward pass usually needs to consume
unaffordable memory to store all the intermediate variables to exactly reverse
the forward training procedure. In this work we propose a simple but effective
method, DrMAD, to distill the knowledge of the forward pass into a shortcut
path, through which we approximately reverse the training trajectory.
Experiments on several image benchmark datasets show that DrMAD is at least 45
times faster and consumes 100 times less memory compared to state-of-the-art
methods for optimizing hyperparameters with minimal compromise to its
effectiveness. To the best of our knowledge, DrMAD is the first research
attempt to make it practical to automatically tune thousands of hyperparameters
of deep neural networks. The code can be downloaded from
https://github.com/bigaidream-projects/drmad
|
Complex Decomposition of the Negative Distance kernel | A Support Vector Machine (SVM) has become a very popular machine learning
method for text classification. One reason for this relates to the range of
existing kernels which allow for classifying data that is not linearly
separable. The linear, polynomial and RBF (Gaussian Radial Basis Function)
kernel are commonly used and serve as a basis of comparison in our study. We
show how to derive the primal form of the quadratic Power Kernel (PK) -- also
called the Negative Euclidean Distance Kernel (NDK) -- by means of complex
numbers. We exemplify the NDK in the framework of text categorization using the
Dewey Document Classification (DDC) as the target scheme. Our evaluation shows
that the power kernel produces F-scores that are comparable to the reference
kernels, but is -- except for the linear kernel -- faster to compute. Finally,
we show how to extend the NDK-approach by including the Mahalanobis distance.
|
Optimally Pruning Decision Tree Ensembles With Feature Cost | We consider the problem of learning decision rules for prediction with
feature budget constraint. In particular, we are interested in pruning an
ensemble of decision trees to reduce expected feature cost while maintaining
high prediction accuracy for any test example. We propose a novel 0-1 integer
program formulation for ensemble pruning. Our pruning formulation is general -
it takes any ensemble of decision trees as input. By explicitly accounting for
feature-sharing across trees together with accuracy/cost trade-off, our method
is able to significantly reduce feature cost by pruning subtrees that introduce
more loss in terms of feature cost than benefit in terms of prediction accuracy
gain. Theoretically, we prove that a linear programming relaxation produces the
exact solution of the original integer program. This allows us to use efficient
convex optimization tools to obtain an optimally pruned ensemble for any given
budget. Empirically, we see that our pruning algorithm significantly improves
the performance of the state of the art ensemble method BudgetRF.
|
A Survey on Social Media Anomaly Detection | Social media anomaly detection is of critical importance to prevent malicious
activities such as bullying, terrorist attack planning, and fraud information
dissemination. With the recent popularity of social media, new types of
anomalous behaviors arise, causing concerns from various parties. While a large
amount of work have been dedicated to traditional anomaly detection problems,
we observe a surge of research interests in the new realm of social media
anomaly detection. In this paper, we present a survey on existing approaches to
address this problem. We focus on the new type of anomalous phenomena in the
social media and review the recent developed techniques to detect those special
types of anomalies. We provide a general overview of the problem domain, common
formulations, existing methodologies and potential directions. With this work,
we hope to call out the attention from the research community on this
challenging problem and open up new directions that we can contribute in the
future.
|
A pragmatic approach to multi-class classification | We present a novel hierarchical approach to multi-class classification which
is generic in that it can be applied to different classification models (e.g.,
support vector machines, perceptrons), and makes no explicit assumptions about
the probabilistic structure of the problem as it is usually done in multi-class
classification. By adding a cascade of additional classifiers, each of which
receives the previous classifier's output in addition to regular input data,
the approach harnesses unused information that manifests itself in the form of,
e.g., correlations between predicted classes. Using multilayer perceptrons as a
classification model, we demonstrate the validity of this approach by testing
it on a complex ten-class 3D gesture recognition task.
|
Streaming Gibbs Sampling for LDA Model | Streaming variational Bayes (SVB) is successful in learning LDA models in an
online manner. However previous attempts toward developing online Monte-Carlo
methods for LDA have little success, often by having much worse perplexity than
their batch counterparts. We present a streaming Gibbs sampling (SGS) method,
an online extension of the collapsed Gibbs sampling (CGS). Our empirical study
shows that SGS can reach similar perplexity as CGS, much better than SVB. Our
distributed version of SGS, DSGS, is much more scalable than SVB mainly because
the updates' communication complexity is small.
|
A simple technique for improving multi-class classification with neural
networks | We present a novel method to perform multi-class pattern classification with
neural networks and test it on a challenging 3D hand gesture recognition
problem. Our method consists of a standard one-against-all (OAA)
classification, followed by another network layer classifying the resulting
class scores, possibly augmented by the original raw input vector. This allows
the network to disambiguate hard-to-separate classes as the distribution of
class scores carries considerable information as well, and is in fact often
used for assessing the confidence of a decision. We show that by this approach
we are able to significantly boost our results, overall as well as for
particular difficult cases, on the hard 10-class gesture classification task.
|
Adaptive and Efficient Nonlinear Channel Equalization for Underwater
Acoustic Communication | We investigate underwater acoustic (UWA) channel equalization and introduce
hierarchical and adaptive nonlinear channel equalization algorithms that are
highly efficient and provide significantly improved bit error rate (BER)
performance. Due to the high complexity of nonlinear equalizers and poor
performance of linear ones, to equalize highly difficult underwater acoustic
channels, we employ piecewise linear equalizers. However, in order to achieve
the performance of the best piecewise linear model, we use a tree structure to
hierarchically partition the space of the received signal. Furthermore, the
equalization algorithm should be completely adaptive, since due to the highly
non-stationary nature of the underwater medium, the optimal MSE equalizer as
well as the best piecewise linear equalizer changes in time. To this end, we
introduce an adaptive piecewise linear equalization algorithm that not only
adapts the linear equalizer at each region but also learns the complete
hierarchical structure with a computational complexity only polynomial in the
number of nodes of the tree. Furthermore, our algorithm is constructed to
directly minimize the final squared error without introducing any ad-hoc
parameters. We demonstrate the performance of our algorithms through highly
realistic experiments performed on accurately simulated underwater acoustic
channels.
|
Angrier Birds: Bayesian reinforcement learning | We train a reinforcement learner to play a simplified version of the game
Angry Birds. The learner is provided with a game state in a manner similar to
the output that could be produced by computer vision algorithms. We improve on
the efficiency of regular {\epsilon}-greedy Q-Learning with linear function
approximation through more systematic exploration in Randomized Least Squares
Value Iteration (RLSVI), an algorithm that samples its policy from a posterior
distribution on optimal policies. With larger state-action spaces, efficient
exploration becomes increasingly important, as evidenced by the faster learning
in RLSVI.
|
From Word Embeddings to Item Recommendation | Social network platforms can use the data produced by their users to serve
them better. One of the services these platforms provide is recommendation
service. Recommendation systems can predict the future preferences of users
using their past preferences. In the recommendation systems literature there
are various techniques, such as neighborhood based methods, machine-learning
based methods and matrix-factorization based methods. In this work, a set of
well known methods from natural language processing domain, namely Word2Vec, is
applied to recommendation systems domain. Unlike previous works that use
Word2Vec for recommendation, this work uses non-textual features, the
check-ins, and it recommends venues to visit/check-in to the target users. For
the experiments, a Foursquare check-in dataset is used. The results show that
use of continuous vector space representations of items modeled by techniques
of Word2Vec is promising for making recommendations.
|
Learning Kernels for Structured Prediction using Polynomial Kernel
Transformations | Learning the kernel functions used in kernel methods has been a vastly
explored area in machine learning. It is now widely accepted that to obtain
'good' performance, learning a kernel function is the key challenge. In this
work we focus on learning kernel representations for structured regression. We
propose use of polynomials expansion of kernels, referred to as Schoenberg
transforms and Gegenbaur transforms, which arise from the seminal result of
Schoenberg (1938). These kernels can be thought of as polynomial combination of
input features in a high dimensional reproducing kernel Hilbert space (RKHS).
We learn kernels over input and output for structured data, such that,
dependency between kernel features is maximized. We use Hilbert-Schmidt
Independence Criterion (HSIC) to measure this. We also give an efficient,
matrix decomposition-based algorithm to learn these kernel transformations, and
demonstrate state-of-the-art results on several real-world datasets.
|
Fast Kronecker product kernel methods via generalized vec trick | Kronecker product kernel provides the standard approach in the kernel methods
literature for learning from graph data, where edges are labeled and both start
and end vertices have their own feature representations. The methods allow
generalization to such new edges, whose start and end vertices do not appear in
the training data, a setting known as zero-shot or zero-data learning. Such a
setting occurs in numerous applications, including drug-target interaction
prediction, collaborative filtering and information retrieval. Efficient
training algorithms based on the so-called vec trick, that makes use of the
special structure of the Kronecker product, are known for the case where the
training data is a complete bipartite graph. In this work we generalize these
results to non-complete training graphs. This allows us to derive a general
framework for training Kronecker product kernel methods, as specific examples
we implement Kronecker ridge regression and support vector machine algorithms.
Experimental results demonstrate that the proposed approach leads to accurate
models, while allowing order of magnitude improvements in training and
prediction time.
|
State Space representation of non-stationary Gaussian Processes | The state space (SS) representation of Gaussian processes (GP) has recently
gained a lot of interest. The main reason is that it allows to compute GPs
based inferences in O(n), where $n$ is the number of observations. This
implementation makes GPs suitable for Big Data. For this reason, it is
important to provide a SS representation of the most important kernels used in
machine learning. The aim of this paper is to show how to exploit the transient
behaviour of SS models to map non-stationary kernels to SS models.
|
An Automaton Learning Approach to Solving Safety Games over Infinite
Graphs | We propose a method to construct finite-state reactive controllers for
systems whose interactions with their adversarial environment are modeled by
infinite-duration two-player games over (possibly) infinite graphs. The
proposed method targets safety games with infinitely many states or with such a
large number of states that it would be impractical---if not impossible---for
conventional synthesis techniques that work on the entire state space. We
resort to constructing finite-state controllers for such systems through an
automata learning approach, utilizing a symbolic representation of the
underlying game that is based on finite automata. Throughout the learning
process, the learner maintains an approximation of the winning region
(represented as a finite automaton) and refines it using different types of
counterexamples provided by the teacher until a satisfactory controller can be
derived (if one exists). We present a symbolic representation of safety games
(inspired by regular model checking), propose implementations of the learner
and teacher, and evaluate their performance on examples motivated by robotic
motion planning in dynamic environments.
|
Ensemble Methods of Classification for Power Systems Security Assessment | One of the most promising approaches for complex technical systems analysis
employs ensemble methods of classification. Ensemble methods enable to build a
reliable decision rules for feature space classification in the presence of
many possible states of the system. In this paper, novel techniques based on
decision trees are used for evaluation of the reliability of the regime of
electric power systems. We proposed hybrid approach based on random forests
models and boosting models. Such techniques can be applied to predict the
interaction of increasing renewable power, storage devices and swiching of
smart loads from intelligent domestic appliances, heaters and air-conditioning
units and electric vehicles with grid for enhanced decision making. The
ensemble classification methods were tested on the modified 118-bus IEEE power
system showing that proposed technique can be employed to examine whether the
power system is secured under steady-state operating conditions.
|
Dense Bag-of-Temporal-SIFT-Words for Time Series Classification | Time series classification is an application of particular interest with the
increase of data to monitor. Classical techniques for time series
classification rely on point-to-point distances. Recently, Bag-of-Words
approaches have been used in this context. Words are quantized versions of
simple features extracted from sliding windows. The SIFT framework has proved
efficient for image classification. In this paper, we design a time series
classification scheme that builds on the SIFT framework adapted to time series
to feed a Bag-of-Words. We then refine our method by studying the impact of
normalized Bag-of-Words, as well as densely extract point descriptors. Proposed
adjustements achieve better performance. The evaluation shows that our method
outperforms classical techniques in terms of classification.
|
Song Recommendation with Non-Negative Matrix Factorization and Graph
Total Variation | This work formulates a novel song recommender system as a matrix completion
problem that benefits from collaborative filtering through Non-negative Matrix
Factorization (NMF) and content-based filtering via total variation (TV) on
graphs. The graphs encode both playlist proximity information and song
similarity, using a rich combination of audio, meta-data and social features.
As we demonstrate, our hybrid recommendation system is very versatile and
incorporates several well-known methods while outperforming them. Particularly,
we show on real-world data that our model overcomes w.r.t. two evaluation
metrics the recommendation of models solely based on low-rank information,
graph-based information or a combination of both.
|
Nonparametric semi-supervised learning of class proportions | The problem of developing binary classifiers from positive and unlabeled data
is often encountered in machine learning. A common requirement in this setting
is to approximate posterior probabilities of positive and negative classes for
a previously unseen data point. This problem can be decomposed into two steps:
(i) the development of accurate predictors that discriminate between positive
and unlabeled data, and (ii) the accurate estimation of the prior probabilities
of positive and negative examples. In this work we primarily focus on the
latter subproblem. We study nonparametric class prior estimation and formulate
this problem as an estimation of mixing proportions in two-component mixture
models, given a sample from one of the components and another sample from the
mixture itself. We show that estimation of mixing proportions is generally
ill-defined and propose a canonical form to obtain identifiability while
maintaining the flexibility to model any distribution. We use insights from
this theory to elucidate the optimization surface of the class priors and
propose an algorithm for estimating them. To address the problems of
high-dimensional density estimation, we provide practical transformations to
low-dimensional spaces that preserve class priors. Finally, we demonstrate the
efficacy of our method on univariate and multivariate data.
|
Scale-Free Online Learning | We design and analyze algorithms for online linear optimization that have
optimal regret and at the same time do not need to know any upper or lower
bounds on the norm of the loss vectors. Our algorithms are instances of the
Follow the Regularized Leader (FTRL) and Mirror Descent (MD) meta-algorithms.
We achieve adaptiveness to the norms of the loss vectors by scale invariance,
i.e., our algorithms make exactly the same decisions if the sequence of loss
vectors is multiplied by any positive constant. The algorithm based on FTRL
works for any decision set, bounded or unbounded. For unbounded decisions sets,
this is the first adaptive algorithm for online linear optimization with a
non-vacuous regret bound. In contrast, we show lower bounds on scale-free
algorithms based on MD on unbounded domains.
|
A note on the sample complexity of the Er-SpUD algorithm by Spielman,
Wang and Wright for exact recovery of sparsely used dictionaries | We consider the problem of recovering an invertible $n \times n$ matrix $A$
and a sparse $n \times p$ random matrix $X$ based on the observation of $Y =
AX$ (up to a scaling and permutation of columns of $A$ and rows of $X$). Using
only elementary tools from the theory of empirical processes we show that a
version of the Er-SpUD algorithm by Spielman, Wang and Wright with high
probability recovers $A$ and $X$ exactly, provided that $p \ge Cn\log n$, which
is optimal up to the constant $C$.
|
On Computationally Tractable Selection of Experiments in
Measurement-Constrained Regression Models | We derive computationally tractable methods to select a small subset of
experiment settings from a large pool of given design points. The primary focus
is on linear regression models, while the technique extends to generalized
linear models and Delta's method (estimating functions of linear regression
models) as well. The algorithms are based on a continuous relaxation of an
otherwise intractable combinatorial optimization problem, with sampling or
greedy procedures as post-processing steps. Formal approximation guarantees are
established for both algorithms, and numerical results on both synthetic and
real-world data confirm the effectiveness of the proposed methods.
|
On Clustering Time Series Using Euclidean Distance and Pearson
Correlation | For time series comparisons, it has often been observed that z-score
normalized Euclidean distances far outperform the unnormalized variant. In this
paper we show that a z-score normalized, squared Euclidean Distance is, in
fact, equal to a distance based on Pearson Correlation. This has profound
impact on many distance-based classification or clustering methods. In addition
to this theoretically sound result we also show that the often used k-Means
algorithm formally needs a mod ification to keep the interpretation as Pearson
correlation strictly valid. Experimental results demonstrate that in many cases
the standard k-Means algorithm generally produces the same results.
|
A Sufficient Statistics Construction of Bayesian Nonparametric
Exponential Family Conjugate Models | Conjugate pairs of distributions over infinite dimensional spaces are
prominent in statistical learning theory, particularly due to the widespread
adoption of Bayesian nonparametric methodologies for a host of models and
applications. Much of the existing literature in the learning community focuses
on processes possessing some form of computationally tractable conjugacy as is
the case for the beta and gamma processes (and, via normalization, the
Dirichlet process). For these processes, proofs of conjugacy and requisite
derivation of explicit computational formulae for posterior density parameters
are idiosyncratic to the stochastic process in question. As such, Bayesian
Nonparametric models are currently available for a limited number of conjugate
pairs, e.g. the Dirichlet-multinomial and beta-Bernoulli process pairs. In each
of these above cases the likelihood process belongs to the class of discrete
exponential family distributions. The exclusion of continuous likelihood
distributions from the known cases of Bayesian Nonparametric Conjugate models
stands as a disparity in the researcher's toolbox.
In this paper we first address the problem of obtaining a general
construction of prior distributions over infinite dimensional spaces possessing
distributional properties amenable to conjugacy. Second, we bridge the divide
between the discrete and continuous likelihoods by illustrating a canonical
construction for stochastic processes whose Levy measure densities are from
positive exponential families, and then demonstrate that these processes in
fact form the prior, likelihood, and posterior in a conjugate family. Our
canonical construction subsumes known computational formulae for posterior
density parameters in the cases where the likelihood is from a discrete
distribution belonging to an exponential family.
|
Temporal Multinomial Mixture for Instance-Oriented Evolutionary
Clustering | Evolutionary clustering aims at capturing the temporal evolution of clusters.
This issue is particularly important in the context of social media data that
are naturally temporally driven. In this paper, we propose a new probabilistic
model-based evolutionary clustering technique. The Temporal Multinomial Mixture
(TMM) is an extension of classical mixture model that optimizes feature
co-occurrences in the trade-off with temporal smoothness. Our model is
evaluated for two recent case studies on opinion aggregation over time. We
compare four different probabilistic clustering models and we show the
superiority of our proposal in the task of instance-oriented clustering.
|
A Synthetic Approach for Recommendation: Combining Ratings, Social
Relations, and Reviews | Recommender systems (RSs) provide an effective way of alleviating the
information overload problem by selecting personalized choices. Online social
networks and user-generated content provide diverse sources for recommendation
beyond ratings, which present opportunities as well as challenges for
traditional RSs. Although social matrix factorization (Social MF) can integrate
ratings with social relations and topic matrix factorization can integrate
ratings with item reviews, both of them ignore some useful information. In this
paper, we investigate the effective data fusion by combining the two
approaches, in two steps. First, we extend Social MF to exploit the graph
structure of neighbors. Second, we propose a novel framework MR3 to jointly
model these three types of information effectively for rating prediction by
aligning latent factors and hidden topics. We achieve more accurate rating
prediction on two real-life datasets. Furthermore, we measure the contribution
of each data source to the proposed framework.
|
Deep Learning over Multi-field Categorical Data: A Case Study on User
Response Prediction | Predicting user responses, such as click-through rate and conversion rate,
are critical in many web applications including web search, personalised
recommendation, and online advertising. Different from continuous raw features
that we usually found in the image and audio domains, the input features in web
space are always of multi-field and are mostly discrete and categorical while
their dependencies are little known. Major user response prediction models have
to either limit themselves to linear models or require manually building up
high-order combination features. The former loses the ability of exploring
feature interactions, while the latter results in a heavy computation in the
large feature space. To tackle the issue, we propose two novel models using
deep neural networks (DNNs) to automatically learn effective patterns from
categorical feature interactions and make predictions of users' ad clicks. To
get our DNNs efficiently work, we propose to leverage three feature
transformation methods, i.e., factorisation machines (FMs), restricted
Boltzmann machines (RBMs) and denoising auto-encoders (DAEs). This paper
presents the structure of our models and their efficient training algorithms.
The large-scale experiments with real-world data demonstrate that our methods
work better than major state-of-the-art models.
|
Implicit Look-alike Modelling in Display Ads: Transfer Collaborative
Filtering to CTR Estimation | User behaviour targeting is essential in online advertising. Compared with
sponsored search keyword targeting and contextual advertising page content
targeting, user behaviour targeting builds users' interest profiles via
tracking their online behaviour and then delivers the relevant ads according to
each user's interest, which leads to higher targeting accuracy and thus more
improved advertising performance. The current user profiling methods include
building keywords and topic tags or mapping users onto a hierarchical taxonomy.
However, to our knowledge, there is no previous work that explicitly
investigates the user online visits similarity and incorporates such similarity
into their ad response prediction. In this work, we propose a general framework
which learns the user profiles based on their online browsing behaviour, and
transfers the learned knowledge onto prediction of their ad response.
Technically, we propose a transfer learning model based on the probabilistic
latent factor graphic models, where the users' ad response profiles are
generated from their online browsing profiles. The large-scale experiments
based on real-world data demonstrate significant improvement of our solution
over some strong baselines.
|
How to learn a graph from smooth signals | We propose a framework that learns the graph structure underlying a set of
smooth signals. Given $X\in\mathbb{R}^{m\times n}$ whose rows reside on the
vertices of an unknown graph, we learn the edge weights
$w\in\mathbb{R}_+^{m(m-1)/2}$ under the smoothness assumption that
$\text{tr}{X^\top LX}$ is small. We show that the problem is a weighted
$\ell$-1 minimization that leads to naturally sparse solutions. We point out
how known graph learning or construction techniques fall within our framework
and propose a new model that performs better than the state of the art in many
settings. We present efficient, scalable primal-dual based algorithms for both
our model and the previous state of the art, and evaluate their performance on
artificial and real data.
|
How to Use Temporal-Driven Constrained Clustering to Detect Typical
Evolutions | In this paper, we propose a new time-aware dissimilarity measure that takes
into account the temporal dimension. Observations that are close in the
description space, but distant in time are considered as dissimilar. We also
propose a method to enforce the segmentation contiguity, by introducing, in the
objective function, a penalty term inspired from the Normal Distribution
Function. We combine the two propositions into a novel time-driven constrained
clustering algorithm, called TDCK-Means, which creates a partition of coherent
clusters, both in the multidimensional space and in the temporal space. This
algorithm uses soft semi-supervised constraints, to encourage adjacent
observations belonging to the same entity to be assigned to the same cluster.
We apply our algorithm to a Political Studies dataset in order to detect
typical evolution phases. We adapt the Shannon entropy in order to measure the
entity contiguity, and we show that our proposition consistently improves
temporal cohesion of clusters, without any significant loss in the
multidimensional variance.
|
Using SVM to pre-classify government purchases | The Brazilian government often misclassifies the goods it buys. That makes it
hard to audit government expenditures. We cannot know whether the price paid
for a ballpoint pen (code #7510) was reasonable if the pen was misclassified as
a technical drawing pen (code #6675) or as any other good. This paper shows how
we can use machine learning to reduce misclassification. I trained a support
vector machine (SVM) classifier that takes a product description as input and
returns the most likely category codes as output. I trained the classifier
using 20 million goods purchased by the Brazilian government between 1999-04-01
and 2015-04-02. In 83.3% of the cases the correct category code was one of the
three most likely category codes identified by the classifier. I used the
trained classifier to develop a web app that might help the government reduce
misclassification. I open sourced the code on GitHub; anyone can use and modify
it.
|
Robobarista: Learning to Manipulate Novel Objects via Deep Multimodal
Embedding | There is a large variety of objects and appliances in human environments,
such as stoves, coffee dispensers, juice extractors, and so on. It is
challenging for a roboticist to program a robot for each of these object types
and for each of their instantiations. In this work, we present a novel approach
to manipulation planning based on the idea that many household objects share
similarly-operated object parts. We formulate the manipulation planning as a
structured prediction problem and learn to transfer manipulation strategy
across different objects by embedding point-cloud, natural language, and
manipulation trajectory data into a shared embedding space using a deep neural
network. In order to learn semantically meaningful spaces throughout our
network, we introduce a method for pre-training its lower layers for multimodal
feature embedding and a method for fine-tuning this embedding space using a
loss-based margin. In order to collect a large number of manipulation
demonstrations for different objects, we develop a new crowd-sourcing platform
called Robobarista. We test our model on our dataset consisting of 116 objects
and appliances with 249 parts along with 250 language instructions, for which
there are 1225 crowd-sourced manipulation demonstrations. We further show that
our robot with our model can even prepare a cup of a latte with appliances it
has never seen before.
|
Deep Learning of Part-based Representation of Data Using Sparse
Autoencoders with Nonnegativity Constraints | We demonstrate a new deep learning autoencoder network, trained by a
nonnegativity constraint algorithm (NCAE), that learns features which show
part-based representation of data. The learning algorithm is based on
constraining negative weights. The performance of the algorithm is assessed
based on decomposing data into parts and its prediction performance is tested
on three standard image data sets and one text dataset. The results indicate
that the nonnegativity constraint forces the autoencoder to learn features that
amount to a part-based representation of data, while improving sparsity and
reconstruction quality in comparison with the traditional sparse autoencoder
and Nonnegative Matrix Factorization. It is also shown that this newly acquired
representation improves the prediction performance of a deep neural network.
|
Learning Hidden Unit Contributions for Unsupervised Acoustic Model
Adaptation | This work presents a broad study on the adaptation of neural network acoustic
models by means of learning hidden unit contributions (LHUC) -- a method that
linearly re-combines hidden units in a speaker- or environment-dependent manner
using small amounts of unsupervised adaptation data. We also extend LHUC to a
speaker adaptive training (SAT) framework that leads to a more adaptable DNN
acoustic model, working both in a speaker-dependent and a speaker-independent
manner, without the requirements to maintain auxiliary speaker-dependent
feature extractors or to introduce significant speaker-dependent changes to the
DNN structure. Through a series of experiments on four different speech
recognition benchmarks (TED talks, Switchboard, AMI meetings, and Aurora4)
comprising 270 test speakers, we show that LHUC in both its test-only and SAT
variants results in consistent word error rate reductions ranging from 5% to
23% relative depending on the task and the degree of mismatch between training
and test data. In addition, we have investigated the effect of the amount of
adaptation data per speaker, the quality of unsupervised adaptation targets,
the complementarity to other adaptation techniques, one-shot adaptation, and an
extension to adapting DNNs trained in a sequence discriminative manner.
|
Online Model Estimation for Predictive Thermal Control of Buildings | This study proposes a general, scalable method to learn control-oriented
thermal models of buildings that could enable wide-scale deployment of
cost-effective predictive controls. An Unscented Kalman Filter augmented for
parameter and disturbance estimation is shown to accurately learn and predict a
building's thermal response. Recent studies of heating, ventilating, and air
conditioning (HVAC) systems have shown significant energy savings with advanced
model predictive control (MPC). A scalable cost-effective method to readily
acquire accurate, robust models of individual buildings' unique thermal
envelopes has historically been elusive and hindered the widespread deployment
of prediction-based control systems. Continuous commissioning and lifetime
performance of these thermal models requires deployment of on-line data-driven
system identification and parameter estimation routines. We propose a novel
gray-box approach using an Unscented Kalman Filter based on a multi-zone
thermal network and validate it with EnergyPlus simulation data. The filter
quickly learns parameters of a thermal network during periods of known or
constrained loads and then characterizes unknown loads in order to provide
accurate 24+ hour energy predictions. This study extends our initial
investigation by formalizing parameter and disturbance estimation routines and
demonstrating results across a year-long study.
|
Infomax strategies for an optimal balance between exploration and
exploitation | Proper balance between exploitation and exploration is what makes good
decisions, which achieve high rewards like payoff or evolutionary fitness. The
Infomax principle postulates that maximization of information directs the
function of diverse systems, from living systems to artificial neural networks.
While specific applications are successful, the validity of information as a
proxy for reward remains unclear. Here, we consider the multi-armed bandit
decision problem, which features arms (slot-machines) of unknown probabilities
of success and a player trying to maximize cumulative payoff by choosing the
sequence of arms to play. We show that an Infomax strategy (Info-p) which
optimally gathers information on the highest mean reward among the arms
saturates known optimal bounds and compares favorably to existing policies. The
highest mean reward considered by Info-p is not the quantity actually needed
for the choice of the arm to play, yet it allows for optimal tradeoffs between
exploration and exploitation.
|
Online Prediction of Dyadic Data with Heterogeneous Matrix Factorization | Dyadic Data Prediction (DDP) is an important problem in many research areas.
This paper develops a novel fully Bayesian nonparametric framework which
integrates two popular and complementary approaches, discrete mixed membership
modeling and continuous latent factor modeling into a unified Heterogeneous
Matrix Factorization~(HeMF) model, which can predict the unobserved dyadics
accurately. The HeMF can determine the number of communities automatically and
exploit the latent linear structure for each bicluster efficiently. We propose
a Variational Bayesian method to estimate the parameters and missing data. We
further develop a novel online learning approach for Variational inference and
use it for the online learning of HeMF, which can efficiently cope with the
important large-scale DDP problem. We evaluate the performance of our method on
the EachMoive, MovieLens and Netflix Prize collaborative filtering datasets.
The experiment shows that, our model outperforms state-of-the-art methods on
all benchmarks. Compared with Stochastic Gradient Method (SGD), our online
learning approach achieves significant improvement on the estimation accuracy
and robustness.
|
Dynamic Privacy For Distributed Machine Learning Over Network | Privacy-preserving distributed machine learning becomes increasingly
important due to the recent rapid growth of data. This paper focuses on a class
of regularized empirical risk minimization (ERM) machine learning problems, and
develops two methods to provide differential privacy to distributed learning
algorithms over a network. We first decentralize the learning algorithm using
the alternating direction method of multipliers (ADMM), and propose the methods
of dual variable perturbation and primal variable perturbation to provide
dynamic differential privacy. The two mechanisms lead to algorithms that can
provide privacy guarantees under mild conditions of the convexity and
differentiability of the loss function and the regularizer. We study the
performance of the algorithms, and show that the dual variable perturbation
outperforms its primal counterpart. To design an optimal privacy mechanisms, we
analyze the fundamental tradeoff between privacy and accuracy, and provide
guidelines to choose privacy parameters. Numerical experiments using customer
information database are performed to corroborate the results on privacy and
utility tradeoffs and design.
|
Deep Learning Applied to Image and Text Matching | The ability to describe images with natural language sentences is the
hallmark for image and language understanding. Such a system has wide ranging
applications such as annotating images and using natural sentences to search
for images.In this project we focus on the task of bidirectional image
retrieval: such asystem is capable of retrieving an image based on a sentence
(image search) andretrieve sentence based on an image query (image annotation).
We present asystem based on a global ranking objective function which uses a
combinationof convolutional neural networks (CNN) and multi layer perceptrons
(MLP).It takes a pair of image and sentence and processes them in different
channels,finally embedding it into a common multimodal vector space. These
embeddingsencode abstract semantic information about the two inputs and can be
comparedusing traditional information retrieval approaches. For each such pair,
the modelreturns a score which is interpretted as a similarity metric. If this
score is high,the image and sentence are likely to convey similar meaning, and
if the score is low then they are likely not to.
The visual input is modeled via deep convolutional neural network. On
theother hand we explore three models for the textual module. The first one
isbag of words with an MLP. The second one uses n-grams (bigram, trigrams,and a
combination of trigram & skip-grams) with an MLP. The third is morespecialized
deep network specific for modeling variable length sequences (SSE).We report
comparable performance to recent work in the field, even though ouroverall
model is simpler. We also show that the training time choice of how wecan
generate our negative samples has a significant impact on performance, and can
be used to specialize the bi-directional system in one particular task.
|
A survey on feature weighting based K-Means algorithms | In a real-world data set there is always the possibility, rather high in our
opinion, that different features may have different degrees of relevance. Most
machine learning algorithms deal with this fact by either selecting or
deselecting features in the data preprocessing phase. However, we maintain that
even among relevant features there may be different degrees of relevance, and
this should be taken into account during the clustering process. With over 50
years of history, K-Means is arguably the most popular partitional clustering
algorithm there is. The first K-Means based clustering algorithm to compute
feature weights was designed just over 30 years ago. Various such algorithms
have been designed since but there has not been, to our knowledge, a survey
integrating empirical evidence of cluster recovery ability, common flaws, and
possible directions for future research. This paper elaborates on the concept
of feature weighting and addresses these issues by critically analysing some of
the most popular, or innovative, feature weighting mechanisms based in K-Means.
|
Creativity in Machine Learning | Recent machine learning techniques can be modified to produce creative
results. Those results did not exist before; it is not a trivial combination of
the data which was fed into the machine learning system. The obtained results
come in multiple forms: As images, as text and as audio.
This paper gives a high level overview of how they are created and gives some
examples. It is meant to be a summary of the current work and give people who
are new to machine learning some starting points.
|
Improved Relation Classification by Deep Recurrent Neural Networks with
Data Augmentation | Nowadays, neural networks play an important role in the task of relation
classification. By designing different neural architectures, researchers have
improved the performance to a large extent in comparison with traditional
methods. However, existing neural networks for relation classification are
usually of shallow architectures (e.g., one-layer convolutional neural networks
or recurrent networks). They may fail to explore the potential representation
space in different abstraction levels. In this paper, we propose deep recurrent
neural networks (DRNNs) for relation classification to tackle this challenge.
Further, we propose a data augmentation method by leveraging the directionality
of relations. We evaluated our DRNNs on the SemEval-2010 Task~8, and achieve an
F1-score of 86.1%, outperforming previous state-of-the-art recorded results.
|
Dual-tree $k$-means with bounded iteration runtime | k-means is a widely used clustering algorithm, but for $k$ clusters and a
dataset size of $N$, each iteration of Lloyd's algorithm costs $O(kN)$ time.
Although there are existing techniques to accelerate single Lloyd iterations,
none of these are tailored to the case of large $k$, which is increasingly
common as dataset sizes grow. We propose a dual-tree algorithm that gives the
exact same results as standard $k$-means; when using cover trees, we use
adaptive analysis techniques to, under some assumptions, bound the
single-iteration runtime of the algorithm as $O(N + k log k)$. To our knowledge
these are the first sub-$O(kN)$ bounds for exact Lloyd iterations. We then show
that this theoretically favorable algorithm performs competitively in practice,
especially for large $N$ and $k$ in low dimensions. Further, the algorithm is
tree-independent, so any type of tree may be used.
|
Linear Algebraic Structure of Word Senses, with Applications to Polysemy | Word embeddings are ubiquitous in NLP and information retrieval, but it is
unclear what they represent when the word is polysemous. Here it is shown that
multiple word senses reside in linear superposition within the word embedding
and simple sparse coding can recover vectors that approximately capture the
senses. The success of our approach, which applies to several embedding
methods, is mathematically explained using a variant of the random walk on
discourses model (Arora et al., 2016). A novel aspect of our technique is that
each extracted word sense is accompanied by one of about 2000 "discourse atoms"
that gives a succinct description of which other words co-occur with that word
sense. Discourse atoms can be of independent interest, and make the method
potentially more useful. Empirical tests are used to verify and support the
theory.
|
Generation of a Supervised Classification Algorithm for Time-Series
Variable Stars with an Application to the LINEAR Dataset | With the advent of digital astronomy, new benefits and new problems have been
presented to the modern day astronomer. While data can be captured in a more
efficient and accurate manor using digital means, the efficiency of data
retrieval has led to an overload of scientific data for processing and storage.
This paper will focus on the construction and application of a supervised
pattern classification algorithm for the identification of variable stars.
Given the reduction of a survey of stars into a standard feature space, the
problem of using prior patterns to identify new observed patterns can be
reduced to time tested classification methodologies and algorithms. Such
supervised methods, so called because the user trains the algorithms prior to
application using patterns with known classes or labels, provide a means to
probabilistically determine the estimated class type of new observations. This
paper will demonstrate the construction and application of a supervised
classification algorithm on variable star data. The classifier is applied to a
set of 192,744 LINEAR data points. Of the original samples, 34,451 unique stars
were classified with high confidence (high level of probability of being the
true class).
|
Trust from the past: Bayesian Personalized Ranking based Link Prediction
in Knowledge Graphs | Link prediction, or predicting the likelihood of a link in a knowledge graph
based on its existing state is a key research task. It differs from a
traditional link prediction task in that the links in a knowledge graph are
categorized into different predicates and the link prediction performance of
different predicates in a knowledge graph generally varies widely. In this
work, we propose a latent feature embedding based link prediction model which
considers the prediction task for each predicate disjointly. To learn the model
parameters it utilizes a Bayesian personalized ranking based optimization
technique. Experimental results on large-scale knowledge bases such as YAGO2
show that our link prediction approach achieves substantially higher
performance than several state-of-art approaches. We also show that for a given
predicate the topological properties of the knowledge graph induced by the
given predicate edges are key indicators of the link prediction performance of
that predicate in the knowledge graph.
|
ActiveClean: Interactive Data Cleaning While Learning Convex Loss Models | Data cleaning is often an important step to ensure that predictive models,
such as regression and classification, are not affected by systematic errors
such as inconsistent, out-of-date, or outlier data. Identifying dirty data is
often a manual and iterative process, and can be challenging on large datasets.
However, many data cleaning workflows can introduce subtle biases into the
training processes due to violation of independence assumptions. We propose
ActiveClean, a progressive cleaning approach where the model is updated
incrementally instead of re-training and can guarantee accuracy on partially
cleaned data. ActiveClean supports a popular class of models called convex loss
models (e.g., linear regression and SVMs). ActiveClean also leverages the
structure of a user's model to prioritize cleaning those records likely to
affect the results. We evaluate ActiveClean on five real-world datasets UCI
Adult, UCI EEG, MNIST, Dollars For Docs, and WorldBank with both real and
synthetic errors. Our results suggest that our proposed optimizations can
improve model accuracy by up-to 2.5x for the same amount of data cleaned.
Furthermore for a fixed cleaning budget and on all real dirty datasets,
ActiveClean returns more accurate models than uniform sampling and Active
Learning.
|
Matrix Neural Networks | Traditional neural networks assume vectorial inputs as the network is
arranged as layers of single line of computing units called neurons. This
special structure requires the non-vectorial inputs such as matrices to be
converted into vectors. This process can be problematic. Firstly, the spatial
information among elements of the data may be lost during vectorisation.
Secondly, the solution space becomes very large which demands very special
treatments to the network parameters and high computational cost. To address
these issues, we propose matrix neural networks (MatNet), which takes matrices
directly as inputs. Each neuron senses summarised information through bilinear
mapping from lower layer units in exactly the same way as the classic feed
forward neural networks. Under this structure, back prorogation and gradient
descent combination can be utilised to obtain network parameters efficiently.
Furthermore, it can be conveniently extended for multimodal inputs. We apply
MatNet to MNIST handwritten digits classification and image super resolution
tasks to show its effectiveness. Without too much tweaking MatNet achieves
comparable performance as the state-of-the-art methods in both tasks with
considerably reduced complexity.
|
On the consistency of inversion-free parameter estimation for Gaussian
random fields | Gaussian random fields are a powerful tool for modeling environmental
processes. For high dimensional samples, classical approaches for estimating
the covariance parameters require highly challenging and massive computations,
such as the evaluation of the Cholesky factorization or solving linear systems.
Recently, Anitescu, Chen and Stein \cite{M.Anitescu} proposed a fast and
scalable algorithm which does not need such burdensome computations. The main
focus of this article is to study the asymptotic behavior of the algorithm of
Anitescu et al. (ACS) for regular and irregular grids in the increasing domain
setting. Consistency, minimax optimality and asymptotic normality of this
algorithm are proved under mild differentiability conditions on the covariance
function. Despite the fact that ACS's method entails a non-concave
maximization, our results hold for any stationary point of the objective
function. A numerical study is presented to evaluate the efficiency of this
algorithm for large data sets.
|
A Relative Exponential Weighing Algorithm for Adversarial Utility-based
Dueling Bandits | We study the K-armed dueling bandit problem which is a variation of the
classical Multi-Armed Bandit (MAB) problem in which the learner receives only
relative feedback about the selected pairs of arms. We propose a new algorithm
called Relative Exponential-weight algorithm for Exploration and Exploitation
(REX3) to handle the adversarial utility-based formulation of this problem.
This algorithm is a non-trivial extension of the Exponential-weight algorithm
for Exploration and Exploitation (EXP3) algorithm. We prove a finite time
expected regret upper bound of order O(sqrt(K ln(K)T)) for this algorithm and a
general lower bound of order omega(sqrt(KT)). At the end, we provide
experimental results using real data from information retrieval applications.
|
Improved graph-based SFA: Information preservation complements the
slowness principle | Slow feature analysis (SFA) is an unsupervised-learning algorithm that
extracts slowly varying features from a multi-dimensional time series. A
supervised extension to SFA for classification and regression is graph-based
SFA (GSFA). GSFA is based on the preservation of similarities, which are
specified by a graph structure derived from the labels. It has been shown that
hierarchical GSFA (HGSFA) allows learning from images and other
high-dimensional data. The feature space spanned by HGSFA is complex due to the
composition of the nonlinearities of the nodes in the network. However, we show
that the network discards useful information prematurely before it reaches
higher nodes, resulting in suboptimal global slowness and an under-exploited
feature space.
To counteract these problems, we propose an extension called hierarchical
information-preserving GSFA (HiGSFA), where information preservation
complements the slowness-maximization goal. We build a 10-layer HiGSFA network
to estimate human age from facial photographs of the MORPH-II database,
achieving a mean absolute error of 3.50 years, improving the state-of-the-art
performance. HiGSFA and HGSFA support multiple-labels and offer a rich feature
space, feed-forward training, and linear complexity in the number of samples
and dimensions. Furthermore, HiGSFA outperforms HGSFA in terms of feature
slowness, estimation accuracy and input reconstruction, giving rise to a
promising hierarchical supervised-learning approach.
|
Average Stability is Invariant to Data Preconditioning. Implications to
Exp-concave Empirical Risk Minimization | We show that the average stability notion introduced by
\cite{kearns1999algorithmic, bousquet2002stability} is invariant to data
preconditioning, for a wide class of generalized linear models that includes
most of the known exp-concave losses. In other words, when analyzing the
stability rate of a given algorithm, we may assume the optimal preconditioning
of the data. This implies that, at least from a statistical perspective,
explicit regularization is not required in order to compensate for
ill-conditioned data, which stands in contrast to a widely common approach that
includes a regularization for analyzing the sample complexity of generalized
linear models. Several important implications of our findings include: a) We
demonstrate that the excess risk of empirical risk minimization (ERM) is
controlled by the preconditioned stability rate. This immediately yields a
relatively short and elegant proof for the fast rates attained by ERM in our
context. b) We strengthen the recent bounds of \cite{hardt2015train} on the
stability rate of the Stochastic Gradient Descent algorithm.
|
Faster Asynchronous SGD | Asynchronous distributed stochastic gradient descent methods have trouble
converging because of stale gradients. A gradient update sent to a parameter
server by a client is stale if the parameters used to calculate that gradient
have since been updated on the server. Approaches have been proposed to
circumvent this problem that quantify staleness in terms of the number of
elapsed updates. In this work, we propose a novel method that quantifies
staleness in terms of moving averages of gradient statistics. We show that this
method outperforms previous methods with respect to convergence speed and
scalability to many clients. We also discuss how an extension to this method
can be used to dramatically reduce bandwidth costs in a distributed training
context. In particular, our method allows reduction of total bandwidth usage by
a factor of 5 with little impact on cost convergence. We also describe (and
link to) a software library that we have used to simulate these algorithms
deterministically on a single machine.
|
Training Recurrent Neural Networks by Diffusion | This work presents a new algorithm for training recurrent neural networks
(although ideas are applicable to feedforward networks as well). The algorithm
is derived from a theory in nonconvex optimization related to the diffusion
equation. The contributions made in this work are two fold. First, we show how
some seemingly disconnected mechanisms used in deep learning such as smart
initialization, annealed learning rate, layerwise pretraining, and noise
injection (as done in dropout and SGD) arise naturally and automatically from
this framework, without manually crafting them into the algorithms. Second, we
present some preliminary results on comparing the proposed method against SGD.
It turns out that the new algorithm can achieve similar level of generalization
accuracy of SGD in much fewer number of epochs.
|
Engineering Safety in Machine Learning | Machine learning algorithms are increasingly influencing our decisions and
interacting with us in all parts of our daily lives. Therefore, just like for
power plants, highways, and myriad other engineered sociotechnical systems, we
must consider the safety of systems involving machine learning. In this paper,
we first discuss the definition of safety in terms of risk, epistemic
uncertainty, and the harm incurred by unwanted outcomes. Then we examine
dimensions, such as the choice of cost function and the appropriateness of
minimizing the empirical average training cost, along which certain real-world
applications may not be completely amenable to the foundational principle of
modern statistical machine learning: empirical risk minimization. In
particular, we note an emerging dichotomy of applications: ones in which safety
is important and risk minimization is not the complete story (we name these
Type A applications), and ones in which safety is not so critical and risk
minimization is sufficient (we name these Type B applications). Finally, we
discuss how four different strategies for achieving safety in engineering
(inherently safe design, safety reserves, safe fail, and procedural safeguards)
can be mapped to the machine learning context through interpretability and
causality of predictive models, objectives beyond expected prediction accuracy,
human involvement for labeling difficult or rare examples, and user experience
design of software.
|
$\mathbf{D^3}$: Deep Dual-Domain Based Fast Restoration of
JPEG-Compressed Images | In this paper, we design a Deep Dual-Domain ($\mathbf{D^3}$) based fast
restoration model to remove artifacts of JPEG compressed images. It leverages
the large learning capacity of deep networks, as well as the problem-specific
expertise that was hardly incorporated in the past design of deep
architectures. For the latter, we take into consideration both the prior
knowledge of the JPEG compression scheme, and the successful practice of the
sparsity-based dual-domain approach. We further design the One-Step Sparse
Inference (1-SI) module, as an efficient and light-weighted feed-forward
approximation of sparse coding. Extensive experiments verify the superiority of
the proposed $D^3$ model over several state-of-the-art methods. Specifically,
our best model is capable of outperforming the latest deep model for around 1
dB in PSNR, and is 30 times faster.
|
Studying Very Low Resolution Recognition Using Deep Networks | Visual recognition research often assumes a sufficient resolution of the
region of interest (ROI). That is usually violated in practice, inspiring us to
explore the Very Low Resolution Recognition (VLRR) problem. Typically, the ROI
in a VLRR problem can be smaller than $16 \times 16$ pixels, and is challenging
to be recognized even by human experts. We attempt to solve the VLRR problem
using deep learning methods. Taking advantage of techniques primarily in super
resolution, domain adaptation and robust regression, we formulate a dedicated
deep learning method and demonstrate how these techniques are incorporated step
by step. Any extra complexity, when introduced, is fully justified by both
analysis and simulation results. The resulting \textit{Robust Partially Coupled
Networks} achieves feature enhancement and recognition simultaneously. It
allows for both the flexibility to combat the LR-HR domain mismatch, and the
robustness to outliers. Finally, the effectiveness of the proposed models is
evaluated on three different VLRR tasks, including face identification, digit
recognition and font recognition, all of which obtain very impressive
performances.
|
Brain-Inspired Deep Networks for Image Aesthetics Assessment | Image aesthetics assessment has been challenging due to its subjective
nature. Inspired by the scientific advances in the human visual perception and
neuroaesthetics, we design Brain-Inspired Deep Networks (BDN) for this task.
BDN first learns attributes through the parallel supervised pathways, on a
variety of selected feature dimensions. A high-level synthesis network is
trained to associate and transform those attributes into the overall aesthetics
rating. We then extend BDN to predicting the distribution of human ratings,
since aesthetics ratings are often subjective. Another highlight is our
first-of-its-kind study of label-preserving transformations in the context of
aesthetics assessment, which leads to an effective data augmentation approach.
Experimental results on the AVA dataset show that our biological inspired and
task-specific BDN model gains significantly performance improvement, compared
to other state-of-the-art models with the same or higher parameter capacity.
|
On-line Bayesian System Identification | We consider an on-line system identification setting, in which new data
become available at given time steps. In order to meet real-time estimation
requirements, we propose a tailored Bayesian system identification procedure,
in which the hyper-parameters are still updated through Marginal Likelihood
maximization, but after only one iteration of a suitable iterative optimization
algorithm. Both gradient methods and the EM algorithm are considered for the
Marginal Likelihood optimization. We compare this "1-step" procedure with the
standard one, in which the optimization method is run until convergence to a
local minimum. The experiments we perform confirm the effectiveness of the
approach we propose.
|
Learning the kernel matrix via predictive low-rank approximations | Efficient and accurate low-rank approximations of multiple data sources are
essential in the era of big data. The scaling of kernel-based learning
algorithms to large datasets is limited by the O(n^2) computation and storage
complexity of the full kernel matrix, which is required by most of the recent
kernel learning algorithms.
We present the Mklaren algorithm to approximate multiple kernel matrices
learn a regression model, which is entirely based on geometrical concepts. The
algorithm does not require access to full kernel matrices yet it accounts for
the correlations between all kernels. It uses Incomplete Cholesky
decomposition, where pivot selection is based on least-angle regression in the
combined, low-dimensional feature space. The algorithm has linear complexity in
the number of data points and kernels. When explicit feature space induced by
the kernel can be constructed, a mapping from the dual to the primal Ridge
regression weights is used for model interpretation.
The Mklaren algorithm was tested on eight standard regression datasets. It
outperforms contemporary kernel matrix approximation approaches when learning
with multiple kernels. It identifies relevant kernels, achieving highest
explained variance than other multiple kernel learning methods for the same
number of iterations. Test accuracy, equivalent to the one using full kernel
matrices, was achieved with at significantly lower approximation ranks. A
difference in run times of two orders of magnitude was observed when either the
number of samples or kernels exceeds 3000.
|
Zero-error dissimilarity based classifiers | We consider general non-Euclidean distance measures between real world
objects that need to be classified. It is assumed that objects are represented
by distances to other objects only. Conditions for zero-error dissimilarity
based classifiers are derived. Additional conditions are given under which the
zero-error decision boundary is a continues function of the distances to a
finite set of training samples. These conditions affect the objects as well as
the distance measure used. It is argued that they can be met in practice.
|
Bandit Structured Prediction for Learning from Partial Feedback in
Statistical Machine Translation | We present an approach to structured prediction from bandit feedback, called
Bandit Structured Prediction, where only the value of a task loss function at a
single predicted point, instead of a correct structure, is observed in
learning. We present an application to discriminative reranking in Statistical
Machine Translation (SMT) where the learning algorithm only has access to a
1-BLEU loss evaluation of a predicted translation instead of obtaining a gold
standard reference translation. In our experiment bandit feedback is obtained
by evaluating BLEU on reference translations without revealing them to the
algorithm. This can be thought of as a simulation of interactive machine
translation where an SMT system is personalized by a user who provides single
point feedback to predicted translations. Our experiments show that our
approach improves translation quality and is comparable to approaches that
employ more informative feedback in learning.
|
Domain based classification | The majority of traditional classification ru les minimizing the expected
probability of error (0-1 loss) are inappropriate if the class probability
distributions are ill-defined or impossible to estimate. We argue that in such
cases class domains should be used instead of class distributions or densities
to construct a reliable decision function. Proposals are presented for some
evaluation criteria and classifier learning schemes, illustrated by an example.
|
Incremental Semiparametric Inverse Dynamics Learning | This paper presents a novel approach for incremental semiparametric inverse
dynamics learning. In particular, we consider the mixture of two approaches:
Parametric modeling based on rigid body dynamics equations and nonparametric
modeling based on incremental kernel methods, with no prior information on the
mechanical properties of the system. This yields to an incremental
semiparametric approach, leveraging the advantages of both the parametric and
nonparametric models. We validate the proposed technique learning the dynamics
of one arm of the iCub humanoid robot.
|
SimpleDS: A Simple Deep Reinforcement Learning Dialogue System | This paper presents 'SimpleDS', a simple and publicly available dialogue
system trained with deep reinforcement learning. In contrast to previous
reinforcement learning dialogue systems, this system avoids manual feature
engineering by performing action selection directly from raw text of the last
system and (noisy) user responses. Our initial results, in the restaurant
domain, show that it is indeed possible to induce reasonable dialogue behaviour
with an approach that aims for high levels of automation in dialogue control
for intelligent interactive agents.
|
Nonparametric Bayesian Storyline Detection from Microtexts | News events and social media are composed of evolving storylines, which
capture public attention for a limited period of time. Identifying storylines
requires integrating temporal and linguistic information, and prior work takes
a largely heuristic approach. We present a novel online non-parametric Bayesian
framework for storyline detection, using the distance-dependent Chinese
Restaurant Process (dd-CRP). To ensure efficient linear-time inference, we
employ a fixed-lag Gibbs sampling procedure, which is novel for the dd-CRP. We
evaluate on the TREC Twitter Timeline Generation (TTG), obtaining encouraging
results: despite using a weak baseline retrieval model, the dd-CRP story
clustering method is competitive with the best entries in the 2014 TTG task.
|
Sparse Convex Clustering | Convex clustering, a convex relaxation of k-means clustering and hierarchical
clustering, has drawn recent attentions since it nicely addresses the
instability issue of traditional nonconvex clustering methods. Although its
computational and statistical properties have been recently studied, the
performance of convex clustering has not yet been investigated in the
high-dimensional clustering scenario, where the data contains a large number of
features and many of them carry no information about the clustering structure.
In this paper, we demonstrate that the performance of convex clustering could
be distorted when the uninformative features are included in the clustering. To
overcome it, we introduce a new clustering method, referred to as Sparse Convex
Clustering, to simultaneously cluster observations and conduct feature
selection. The key idea is to formulate convex clustering in a form of
regularization, with an adaptive group-lasso penalty term on cluster centers.
In order to optimally balance the tradeoff between the cluster fitting and
sparsity, a tuning criterion based on clustering stability is developed. In
theory, we provide an unbiased estimator for the degrees of freedom of the
proposed sparse convex clustering method. Finally, the effectiveness of the
sparse convex clustering is examined through a variety of numerical experiments
and a real data application.
|
Spectral Theory of Unsigned and Signed Graphs. Applications to Graph
Clustering: a Survey | This is a survey of the method of graph cuts and its applications to graph
clustering of weighted unsigned and signed graphs. I provide a fairly thorough
treatment of the method of normalized graph cuts, a deeply original method due
to Shi and Malik, including complete proofs. The main thrust of this paper is
the method of normalized cuts. I give a detailed account for K = 2 clusters,
and also for K > 2 clusters, based on the work of Yu and Shi. I also show how
both graph drawing and normalized cut K-clustering can be easily generalized to
handle signed graphs, which are weighted graphs in which the weight matrix W
may have negative coefficients. Intuitively, negative coefficients indicate
distance or dissimilarity. The solution is to replace the degree matrix by the
matrix in which absolute values of the weights are used, and to replace the
Laplacian by the Laplacian with the new degree matrix of absolute values. As
far as I know, the generalization of K-way normalized clustering to signed
graphs is new. Finally, I show how the method of ratio cuts, in which a cut is
normalized by the size of the cluster rather than its volume, is just a special
case of normalized cuts.
|
Sub-Sampled Newton Methods I: Globally Convergent Algorithms | Large scale optimization problems are ubiquitous in machine learning and data
analysis and there is a plethora of algorithms for solving such problems. Many
of these algorithms employ sub-sampling, as a way to either speed up the
computations and/or to implicitly implement a form of statistical
regularization. In this paper, we consider second-order iterative optimization
algorithms and we provide bounds on the convergence of the variants of Newton's
method that incorporate uniform sub-sampling as a means to estimate the
gradient and/or Hessian. Our bounds are non-asymptotic and quantitative. Our
algorithms are global and are guaranteed to converge from any initial iterate.
Using random matrix concentration inequalities, one can sub-sample the
Hessian to preserve the curvature information. Our first algorithm incorporates
Hessian sub-sampling while using the full gradient. We also give additional
convergence results for when the sub-sampled Hessian is regularized by
modifying its spectrum or ridge-type regularization. Next, in addition to
Hessian sub-sampling, we also consider sub-sampling the gradient as a way to
further reduce the computational complexity per iteration. We use approximate
matrix multiplication results from randomized numerical linear algebra to
obtain the proper sampling strategy. In all these algorithms, computing the
update boils down to solving a large scale linear system, which can be
computationally expensive. As a remedy, for all of our algorithms, we also give
global convergence results for the case of inexact updates where such linear
system is solved only approximately.
This paper has a more advanced companion paper, [42], in which we demonstrate
that, by doing a finer-grained analysis, we can get problem-independent bounds
for local convergence of these algorithms and explore trade-offs to improve
upon the basic results of the present paper.
|
Sub-Sampled Newton Methods II: Local Convergence Rates | Many data-fitting applications require the solution of an optimization
problem involving a sum of large number of functions of high dimensional
parameter. Here, we consider the problem of minimizing a sum of $n$ functions
over a convex constraint set $\mathcal{X} \subseteq \mathbb{R}^{p}$ where both
$n$ and $p$ are large. In such problems, sub-sampling as a way to reduce $n$
can offer great amount of computational efficiency.
Within the context of second order methods, we first give quantitative local
convergence results for variants of Newton's method where the Hessian is
uniformly sub-sampled. Using random matrix concentration inequalities, one can
sub-sample in a way that the curvature information is preserved. Using such
sub-sampling strategy, we establish locally Q-linear and Q-superlinear
convergence rates. We also give additional convergence results for when the
sub-sampled Hessian is regularized by modifying its spectrum or Levenberg-type
regularization.
Finally, in addition to Hessian sub-sampling, we consider sub-sampling the
gradient as way to further reduce the computational complexity per iteration.
We use approximate matrix multiplication results from randomized numerical
linear algebra (RandNLA) to obtain the proper sampling strategy and we
establish locally R-linear convergence rates. In such a setting, we also show
that a very aggressive sample size increase results in a R-superlinearly
convergent algorithm.
While the sample size depends on the condition number of the problem, our
convergence rates are problem-independent, i.e., they do not depend on the
quantities related to the problem. Hence, our analysis here can be used to
complement the results of our basic framework from the companion paper, [38],
by exploring algorithmic trade-offs that are important in practice.
|
Improved Sampling Techniques for Learning an Imbalanced Data Set | This paper presents the performance of a classifier built using the stackingC
algorithm in nine different data sets. Each data set is generated using a
sampling technique applied on the original imbalanced data set. Five new
sampling techniques are proposed in this paper (i.e., SMOTERandRep, Lax Random
Oversampling, Lax Random Undersampling, Combined-Lax Random Oversampling
Undersampling, and Combined-Lax Random Undersampling Oversampling) that were
based on the three sampling techniques (i.e., Random Undersampling, Random
Oversampling, and Synthetic Minority Oversampling Technique) usually used as
solutions in imbalance learning. The metrics used to evaluate the classifier's
performance were F-measure and G-mean. F-measure determines the performance of
the classifier for every class, while G-mean measures the overall performance
of the classifier. The results using F-measure showed that for the data without
a sampling technique, the classifier's performance is good only for the
majority class. It also showed that among the eight sampling techniques, RU and
LRU have the worst performance while other techniques (i.e., RO, C-LRUO and
C-LROU) performed well only on some classes. The best performing techniques in
all data sets were SMOTE, SMOTERandRep, and LRO having the lowest F-measure
values between 0.5 and 0.65. The results using G-mean showed that the
oversampling technique that attained the highest G-mean value is LRO (0.86),
next is C-LROU (0.85), then SMOTE (0.84) and finally is SMOTERandRep (0.83).
Combining the result of the two metrics (F-measure and G-mean), only the three
sampling techniques are considered as good performing (i.e., LRO, SMOTE, and
SMOTERandRep).
|
Top-N Recommender System via Matrix Completion | Top-N recommender systems have been investigated widely both in industry and
academia. However, the recommendation quality is far from satisfactory. In this
paper, we propose a simple yet promising algorithm. We fill the user-item
matrix based on a low-rank assumption and simultaneously keep the original
information. To do that, a nonconvex rank relaxation rather than the nuclear
norm is adopted to provide a better rank approximation and an efficient
optimization strategy is designed. A comprehensive set of experiments on real
datasets demonstrates that our method pushes the accuracy of Top-N
recommendation to a new level.
|
Understanding Deep Convolutional Networks | Deep convolutional networks provide state of the art classifications and
regressions results over many high-dimensional problems. We review their
architecture, which scatters data with a cascade of linear filter weights and
non-linearities. A mathematical framework is introduced to analyze their
properties. Computations of invariants involve multiscale contractions, the
linearization of hierarchical symmetries, and sparse separations. Applications
are discussed.
|
A Theory of Local Matching: SIFT and Beyond | Why has SIFT been so successful? Why its extension, DSP-SIFT, can further
improve SIFT? Is there a theory that can explain both? How can such theory
benefit real applications? Can it suggest new algorithms with reduced
computational complexity or new descriptors with better accuracy for matching?
We construct a general theory of local descriptors for visual matching. Our
theory relies on concepts in energy minimization and heat diffusion. We show
that SIFT and DSP-SIFT approximate the solution the theory suggests. In
particular, DSP-SIFT gives a better approximation to the theoretical solution;
justifying why DSP-SIFT outperforms SIFT. Using the developed theory, we derive
new descriptors that have fewer parameters and are potentially better in
handling affine deformations.
|
Habits vs Environment: What really causes asthma? | Despite considerable number of studies on risk factors for asthma onset, very
little is known about their relative importance. To have a full picture of
these factors, both categories, personal and environmental data, have to be
taken into account simultaneously, which is missing in previous studies. We
propose a framework to rank the risk factors from heterogeneous data sources of
the two categories. Established on top of EventShop and Personal EventShop,
this framework extracts about 400 features, and analyzes them by employing a
gradient boosting tree. The features come from sources including personal
profile and life-event data, and environmental data on air pollution, weather
and PM2.5 emission sources. The top ranked risk factors derived from our
framework agree well with the general medical consensus. Thus, our framework is
a reliable approach, and the discovered rankings of relative importance of risk
factors can provide insights for the prevention of asthma.
|
Data-driven Rank Breaking for Efficient Rank Aggregation | Rank aggregation systems collect ordinal preferences from individuals to
produce a global ranking that represents the social preference. Rank-breaking
is a common practice to reduce the computational complexity of learning the
global ranking. The individual preferences are broken into pairwise comparisons
and applied to efficient algorithms tailored for independent paired
comparisons. However, due to the ignored dependencies in the data, naive
rank-breaking approaches can result in inconsistent estimates. The key idea to
produce accurate and consistent estimates is to treat the pairwise comparisons
unequally, depending on the topology of the collected data. In this paper, we
provide the optimal rank-breaking estimator, which not only achieves
consistency but also achieves the best error bound. This allows us to
characterize the fundamental tradeoff between accuracy and complexity. Further,
the analysis identifies how the accuracy depends on the spectral gap of a
corresponding comparison graph.
|
Incremental Spectral Sparsification for Large-Scale Graph-Based
Semi-Supervised Learning | While the harmonic function solution performs well in many semi-supervised
learning (SSL) tasks, it is known to scale poorly with the number of samples.
Recent successful and scalable methods, such as the eigenfunction method focus
on efficiently approximating the whole spectrum of the graph Laplacian
constructed from the data. This is in contrast to various subsampling and
quantization methods proposed in the past, which may fail in preserving the
graph spectra. However, the impact of the approximation of the spectrum on the
final generalization error is either unknown, or requires strong assumptions on
the data. In this paper, we introduce Sparse-HFS, an efficient
edge-sparsification algorithm for SSL. By constructing an edge-sparse and
spectrally similar graph, we are able to leverage the approximation guarantees
of spectral sparsification methods to bound the generalization error of
Sparse-HFS. As a result, we obtain a theoretically-grounded approximation
scheme for graph-based SSL that also empirically matches the performance of
known large-scale methods.
|
A Confidence-Based Approach for Balancing Fairness and Accuracy | We study three classical machine learning algorithms in the context of
algorithmic fairness: adaptive boosting, support vector machines, and logistic
regression. Our goal is to maintain the high accuracy of these learning
algorithms while reducing the degree to which they discriminate against
individuals because of their membership in a protected group.
Our first contribution is a method for achieving fairness by shifting the
decision boundary for the protected group. The method is based on the theory of
margins for boosting. Our method performs comparably to or outperforms previous
algorithms in the fairness literature in terms of accuracy and low
discrimination, while simultaneously allowing for a fast and transparent
quantification of the trade-off between bias and error.
Our second contribution addresses the shortcomings of the bias-error
trade-off studied in most of the algorithmic fairness literature. We
demonstrate that even hopelessly naive modifications of a biased algorithm,
which cannot be reasonably said to be fair, can still achieve low bias and high
accuracy. To help to distinguish between these naive algorithms and more
sensible algorithms we propose a new measure of fairness, called resilience to
random bias (RRB). We demonstrate that RRB distinguishes well between our naive
and sensible fairness algorithms. RRB together with bias and accuracy provides
a more complete picture of the fairness of an algorithm.
|
Local Network Community Detection with Continuous Optimization of
Conductance and Weighted Kernel K-Means | Local network community detection is the task of finding a single community
of nodes concentrated around few given seed nodes in a localized way.
Conductance is a popular objective function used in many algorithms for local
community detection. This paper studies a continuous relaxation of conductance.
We show that continuous optimization of this objective still leads to discrete
communities. We investigate the relation of conductance with weighted kernel
k-means for a single community, which leads to the introduction of a new
objective function, $\sigma$-conductance. Conductance is obtained by setting
$\sigma$ to $0$. Two algorithms, EMc and PGDc, are proposed to locally optimize
$\sigma$-conductance and automatically tune the parameter $\sigma$. They are
based on expectation maximization and projected gradient descent, respectively.
We prove locality and give performance guarantees for EMc and PGDc for a class
of dense and well separated communities centered around the seeds. Experiments
are conducted on networks with ground-truth communities, comparing to
state-of-the-art graph diffusion algorithms for conductance optimization. On
large graphs, results indicate that EMc and PGDc stay localized and produce
communities most similar to the ground, while graph diffusion algorithms
generate large communities of lower quality.
|
When is Clustering Perturbation Robust? | Clustering is a fundamental data mining tool that aims to divide data into
groups of similar items. Generally, intuition about clustering reflects the
ideal case -- exact data sets endowed with flawless dissimilarity between
individual instances.
In practice however, these cases are in the minority, and clustering
applications are typically characterized by noisy data sets with approximate
pairwise dissimilarities. As such, the efficacy of clustering methods in
practical applications necessitates robustness to perturbations.
In this paper, we perform a formal analysis of perturbation robustness,
revealing that the extent to which algorithms can exhibit this desirable
characteristic is inherently limited, and identifying the types of structures
that allow popular clustering paradigms to discover meaningful clusters in
spite of faulty data.
|
Exploiting Low-dimensional Structures to Enhance DNN Based Acoustic
Modeling in Speech Recognition | We propose to model the acoustic space of deep neural network (DNN)
class-conditional posterior probabilities as a union of low-dimensional
subspaces. To that end, the training posteriors are used for dictionary
learning and sparse coding. Sparse representation of the test posteriors using
this dictionary enables projection to the space of training data. Relying on
the fact that the intrinsic dimensions of the posterior subspaces are indeed
very small and the matrix of all posteriors belonging to a class has a very low
rank, we demonstrate how low-dimensional structures enable further enhancement
of the posteriors and rectify the spurious errors due to mismatch conditions.
The enhanced acoustic modeling method leads to improvements in continuous
speech recognition task using hybrid DNN-HMM (hidden Markov model) framework in
both clean and noisy conditions, where upto 15.4% relative reduction in word
error rate (WER) is achieved.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.