title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Learning Adversary-Resistant Deep Neural Networks | Deep neural networks (DNNs) have proven to be quite effective in a vast array
of machine learning tasks, with recent examples in cyber security and
autonomous vehicles. Despite the superior performance of DNNs in these
applications, it has been recently shown that these models are susceptible to a
particular type of attack that exploits a fundamental flaw in their design.
This attack consists of generating particular synthetic examples referred to as
adversarial samples. These samples are constructed by slightly manipulating
real data-points in order to "fool" the original DNN model, forcing it to
mis-classify previously correctly classified samples with high confidence.
Addressing this flaw in the model is essential if DNNs are to be used in
critical applications such as those in cyber security.
Previous work has provided various learning algorithms to enhance the
robustness of DNN models, and they all fall into the tactic of "security
through obscurity". This means security can be guaranteed only if one can
obscure the learning algorithms from adversaries. Once the learning technique
is disclosed, DNNs protected by these defense mechanisms are still susceptible
to adversarial samples. In this work, we investigate this issue shared across
previous research work and propose a generic approach to escalate a DNN's
resistance to adversarial samples. More specifically, our approach integrates a
data transformation module with a DNN, making it robust even if we reveal the
underlying learning algorithm. To demonstrate the generality of our proposed
approach and its potential for handling cyber security applications, we
evaluate our method and several other existing solutions on datasets publicly
available. Our results indicate that our approach typically provides superior
classification performance and resistance in comparison with state-of-art
solutions.
|
Semi-Supervised Learning via Sparse Label Propagation | This work proposes a novel method for semi-supervised learning from partially
labeled massive network-structured datasets, i.e., big data over networks. We
model the underlying hypothesis, which relates data points to labels, as a
graph signal, defined over some graph (network) structure intrinsic to the
dataset. Following the key principle of supervised learning, i.e., similar
inputs yield similar outputs, we require the graph signals induced by labels to
have small total variation. Accordingly, we formulate the problem of learning
the labels of data points as a non-smooth convex optimization problem which
amounts to balancing between the empirical loss, i.e., the discrepancy with
some partially available label information, and the smoothness quantified by
the total variation of the learned graph signal. We solve this optimization
problem by appealing to a recently proposed preconditioned variant of the
popular primal-dual method by Pock and Chambolle, which results in a sparse
label propagation algorithm. This learning algorithm allows for a highly
scalable implementation as message passing over the underlying data graph. By
applying concepts of compressed sensing to the learning problem, we are also
able to provide a transparent sufficient condition on the underlying network
structure such that accurate learning of the labels is possible. We also
present an implementation of the message passing formulation allows for a
highly scalable implementation in big data frameworks.
|
Zeroth-order Asynchronous Doubly Stochastic Algorithm with Variance
Reduction | Zeroth-order (derivative-free) optimization attracts a lot of attention in
machine learning, because explicit gradient calculations may be computationally
expensive or infeasible. To handle large scale problems both in volume and
dimension, recently asynchronous doubly stochastic zeroth-order algorithms were
proposed. The convergence rate of existing asynchronous doubly stochastic
zeroth order algorithms is $O(\frac{1}{\sqrt{T}})$ (also for the sequential
stochastic zeroth-order optimization algorithms). In this paper, we focus on
the finite sums of smooth but not necessarily convex functions, and propose an
asynchronous doubly stochastic zeroth-order optimization algorithm using the
accelerated technology of variance reduction (AsyDSZOVR). Rigorous theoretical
analysis show that the convergence rate can be improved from
$O(\frac{1}{\sqrt{T}})$ the best result of existing algorithms to
$O(\frac{1}{T})$. Also our theoretical results is an improvement to the ones of
the sequential stochastic zeroth-order optimization algorithms.
|
Extracting Implicit Social Relation for Social Recommendation Techniques
in User Rating Prediction | Recommendation plays an increasingly important role in our daily lives.
Recommender systems automatically suggest items to users that might be
interesting for them. Recent studies illustrate that incorporating social trust
in Matrix Factorization methods demonstrably improves accuracy of rating
prediction. Such approaches mainly use the trust scores explicitly expressed by
users. However, it is often challenging to have users provide explicit trust
scores of each other. There exist quite a few works, which propose Trust
Metrics to compute and predict trust scores between users based on their
interactions. In this paper, first we present how social relation can be
extracted from users' ratings to items by describing Hellinger distance between
users in recommender systems. Then, we propose to incorporate the predicted
trust scores into social matrix factorization models. By analyzing social
relation extraction from three well-known real-world datasets, which both:
trust and recommendation data available, we conclude that using the implicit
social relation in social recommendation techniques has almost the same
performance compared to the actual trust scores explicitly expressed by users.
Hence, we build our method, called Hell-TrustSVD, on top of the
state-of-the-art social recommendation technique to incorporate both the
extracted implicit social relations and ratings given by users on the
prediction of items for an active user. To the best of our knowledge, this is
the first work to extend TrustSVD with extracted social trust information. The
experimental results support the idea of employing implicit trust into matrix
factorization whenever explicit trust is not available, can perform much better
than the state-of-the-art approaches in user rating prediction.
|
Understanding and Optimizing the Performance of Distributed Machine
Learning Applications on Apache Spark | In this paper we explore the performance limits of Apache Spark for machine
learning applications. We begin by analyzing the characteristics of a
state-of-the-art distributed machine learning algorithm implemented in Spark
and compare it to an equivalent reference implementation using the high
performance computing framework MPI. We identify critical bottlenecks of the
Spark framework and carefully study their implications on the performance of
the algorithm. In order to improve Spark performance we then propose a number
of practical techniques to alleviate some of its overheads. However, optimizing
computational efficiency and framework related overheads is not the only key to
performance -- we demonstrate that in order to get the best performance out of
any implementation it is necessary to carefully tune the algorithm to the
respective trade-off between computation time and communication latency. The
optimal trade-off depends on both the properties of the distributed algorithm
as well as infrastructure and framework-related characteristics. Finally, we
apply these technical and algorithmic optimizations to three different
distributed linear machine learning algorithms that have been implemented in
Spark. We present results using five large datasets and demonstrate that by
using the proposed optimizations, we can achieve a reduction in the performance
difference between Spark and MPI from 20x to 2x.
|
Support vector regression model for BigData systems | Nowadays Big Data are becoming more and more important. Many sectors of our
economy are now guided by data-driven decision processes. Big Data and business
intelligence applications are facilitated by the MapReduce programming model
while, at infrastructural layer, cloud computing provides flexible and cost
effective solutions for allocating on demand large clusters. In such systems,
capacity allocation, which is the ability to optimally size minimal resources
for achieve a certain level of performance, is a key challenge to enhance
performance for MapReduce jobs and minimize cloud resource costs. In order to
do so, one of the biggest challenge is to build an accurate performance model
to estimate job execution time of MapReduce systems. Previous works applied
simulation based models for modeling such systems. Although this approach can
accurately describe the behavior of Big Data clusters, it is too
computationally expensive and does not scale to large system. We try to
overcome these issues by applying machine learning techniques. More precisely
we focus on Support Vector Regression (SVR) which is intrinsically more robust
w.r.t other techniques, like, e.g., neural networks, and less sensitive to
outliers in the training set. To better investigate these benefits, we compare
SVR to linear regression.
|
Simple and Scalable Predictive Uncertainty Estimation using Deep
Ensembles | Deep neural networks (NNs) are powerful black box predictors that have
recently achieved impressive performance on a wide spectrum of tasks.
Quantifying predictive uncertainty in NNs is a challenging and yet unsolved
problem. Bayesian NNs, which learn a distribution over weights, are currently
the state-of-the-art for estimating predictive uncertainty; however these
require significant modifications to the training procedure and are
computationally expensive compared to standard (non-Bayesian) NNs. We propose
an alternative to Bayesian NNs that is simple to implement, readily
parallelizable, requires very little hyperparameter tuning, and yields high
quality predictive uncertainty estimates. Through a series of experiments on
classification and regression benchmarks, we demonstrate that our method
produces well-calibrated uncertainty estimates which are as good or better than
approximate Bayesian NNs. To assess robustness to dataset shift, we evaluate
the predictive uncertainty on test examples from known and unknown
distributions, and show that our method is able to express higher uncertainty
on out-of-distribution examples. We demonstrate the scalability of our method
by evaluating predictive uncertainty estimates on ImageNet.
|
Generalized RBF kernel for incomplete data | We construct $\bf genRBF$ kernel, which generalizes the classical Gaussian
RBF kernel to the case of incomplete data. We model the uncertainty contained
in missing attributes making use of data distribution and associate every point
with a conditional probability density function. This allows to embed
incomplete data into the function space and to define a kernel between two
missing data points based on scalar product in $L_2$. Experiments show that
introduced kernel applied to SVM classifier gives better results than other
state-of-the-art methods, especially in the case when large number of features
is missing. Moreover, it is easy to implement and can be used together with any
kernel approaches with no additional modifications.
|
A Nonparametric Latent Factor Model For Location-Aware Video
Recommendations | We are interested in learning customers' video preferences from their
historic viewing patterns and geographical location. We consider a Bayesian
latent factor modeling approach for this task. In order to tune the complexity
of the model to best represent the data, we make use of Bayesian nonparameteric
techniques. We describe an inference technique that can scale to large
real-world data sets. Finally we show results obtained by applying the model to
a large internal Netflix data set, that illustrates that the model was able to
capture interesting relationships between viewing patterns and geographical
location.
|
Towards the Limit of Network Quantization | Network quantization is one of network compression techniques to reduce the
redundancy of deep neural networks. It reduces the number of distinct network
parameter values by quantization in order to save the storage for them. In this
paper, we design network quantization schemes that minimize the performance
loss due to quantization given a compression ratio constraint. We analyze the
quantitative relation of quantization errors to the neural network loss
function and identify that the Hessian-weighted distortion measure is locally
the right objective function for the optimization of network quantization. As a
result, Hessian-weighted k-means clustering is proposed for clustering network
parameters to quantize. When optimal variable-length binary codes, e.g.,
Huffman codes, are employed for further compression, we derive that the network
quantization problem can be related to the entropy-constrained scalar
quantization (ECSQ) problem in information theory and consequently propose two
solutions of ECSQ for network quantization, i.e., uniform quantization and an
iterative solution similar to Lloyd's algorithm. Finally, using the simple
uniform quantization followed by Huffman coding, we show from our experiments
that the compression ratios of 51.25, 22.17 and 40.65 are achievable for LeNet,
32-layer ResNet and AlexNet, respectively.
|
Improving the Performance of Neural Networks in Regression Tasks Using
Drawering | The method presented extends a given regression neural network to make its
performance improve. The modification affects the learning procedure only,
hence the extension may be easily omitted during evaluation without any change
in prediction. It means that the modified model may be evaluated as quickly as
the original one but tends to perform better.
This improvement is possible because the modification gives better expressive
power, provides better behaved gradients and works as a regularization. The
knowledge gained by the temporarily extended neural network is contained in the
parameters shared with the original neural network.
The only cost is an increase in learning time.
|
Deterministic and Probabilistic Conditions for Finite Completability of
Low-Tucker-Rank Tensor | We investigate the fundamental conditions on the sampling pattern, i.e.,
locations of the sampled entries, for finite completability of a low-rank
tensor given some components of its Tucker rank. In order to find the
deterministic necessary and sufficient conditions, we propose an algebraic
geometric analysis on the Tucker manifold, which allows us to incorporate
multiple rank components in the proposed analysis in contrast with the
conventional geometric approaches on the Grassmannian manifold. This analysis
characterizes the algebraic independence of a set of polynomials defined based
on the sampling pattern, which is closely related to finite completion.
Probabilistic conditions are then studied and a lower bound on the sampling
probability is given, which guarantees that the proposed deterministic
conditions on the sampling patterns for finite completability hold with high
probability. Furthermore, using the proposed geometric approach for finite
completability, we propose a sufficient condition on the sampling pattern that
ensures there exists exactly one completion for the sampled tensor.
|
Distributed Gaussian Learning over Time-varying Directed Graphs | We present a distributed (non-Bayesian) learning algorithm for the problem of
parameter estimation with Gaussian noise. The algorithm is expressed as
explicit updates on the parameters of the Gaussian beliefs (i.e. means and
precision). We show a convergence rate of $O(1/k)$ with the constant term
depending on the number of agents and the topology of the network. Moreover, we
show almost sure convergence to the optimal solution of the estimation problem
for the general case of time-varying directed graphs.
|
Efficient Non-oblivious Randomized Reduction for Risk Minimization with
Improved Excess Risk Guarantee | In this paper, we address learning problems for high dimensional data.
Previously, oblivious random projection based approaches that project high
dimensional features onto a random subspace have been used in practice for
tackling high-dimensionality challenge in machine learning. Recently, various
non-oblivious randomized reduction methods have been developed and deployed for
solving many numerical problems such as matrix product approximation, low-rank
matrix approximation, etc. However, they are less explored for the machine
learning tasks, e.g., classification. More seriously, the theoretical analysis
of excess risk bounds for risk minimization, an important measure of
generalization performance, has not been established for non-oblivious
randomized reduction methods. It therefore remains an open problem what is the
benefit of using them over previous oblivious random projection based
approaches. To tackle these challenges, we propose an algorithmic framework for
employing non-oblivious randomized reduction method for general empirical risk
minimizing in machine learning tasks, where the original high-dimensional
features are projected onto a random subspace that is derived from the data
with a small matrix approximation error. We then derive the first excess risk
bound for the proposed non-oblivious randomized reduction approach without
requiring strong assumptions on the training data. The established excess risk
bound exhibits that the proposed approach provides much better generalization
performance and it also sheds more insights about different randomized
reduction approaches. Finally, we conduct extensive experiments on both
synthetic and real-world benchmark datasets, whose dimension scales to
$O(10^7)$, to demonstrate the efficacy of our proposed approach.
|
Statistical mechanics of unsupervised feature learning in a restricted
Boltzmann machine with binary synapses | Revealing hidden features in unlabeled data is called unsupervised feature
learning, which plays an important role in pretraining a deep neural network.
Here we provide a statistical mechanics analysis of the unsupervised learning
in a restricted Boltzmann machine with binary synapses. A message passing
equation to infer the hidden feature is derived, and furthermore, variants of
this equation are analyzed. A statistical analysis by replica theory describes
the thermodynamic properties of the model. Our analysis confirms an entropy
crisis preceding the non-convergence of the message passing equation,
suggesting a discontinuous phase transition as a key characteristic of the
restricted Boltzmann machine. Continuous phase transition is also confirmed
depending on the embedded feature strength in the data. The mean-field result
under the replica symmetric assumption agrees with that obtained by running
message passing algorithms on single instances of finite sizes. Interestingly,
in an approximate Hopfield model, the entropy crisis is absent, and a
continuous phase transition is observed instead. We also develop an iterative
equation to infer the hyper-parameter (temperature) hidden in the data, which
in physics corresponds to iteratively imposing Nishimori condition. Our study
provides insights towards understanding the thermodynamic properties of the
restricted Boltzmann machine learning, and moreover important theoretical basis
to build simplified deep networks.
|
Factored Contextual Policy Search with Bayesian Optimization | Scarce data is a major challenge to scaling robot learning to truly complex
tasks, as we need to generalize locally learned policies over different
"contexts". Bayesian optimization approaches to contextual policy search (CPS)
offer data-efficient policy learning that generalize over a context space. We
propose to improve data-efficiency by factoring typically considered contexts
into two components: target-type contexts that correspond to a desired outcome
of the learned behavior, e.g. target position for throwing a ball; and
environment type contexts that correspond to some state of the environment,
e.g. initial ball position or wind speed. Our key observation is that
experience can be directly generalized over target-type contexts. Based on that
we introduce Factored Contextual Policy Search with Bayesian Optimization for
both passive and active learning settings. Preliminary results show faster
policy generalization on a simulated toy problem. A full paper extension is
available at arXiv:1904.11761
|
Video Ladder Networks | We present the Video Ladder Network (VLN) for efficiently generating future
video frames. VLN is a neural encoder-decoder model augmented at all layers by
both recurrent and feedforward lateral connections. At each layer, these
connections form a lateral recurrent residual block, where the feedforward
connection represents a skip connection and the recurrent connection represents
the residual. Thanks to the recurrent connections, the decoder can exploit
temporal summaries generated from all layers of the encoder. This way, the top
layer is relieved from the pressure of modeling lower-level spatial and
temporal details. Furthermore, we extend the basic version of VLN to
incorporate ResNet-style residual blocks in the encoder and decoder, which help
improving the prediction results. VLN is trained in self-supervised regime on
the Moving MNIST dataset, achieving competitive results while having very
simple structure and providing fast inference.
|
Control Matching via Discharge Code Sequences | In this paper, we consider the patient similarity matching problem over a
cancer cohort of more than 220,000 patients. Our approach first leverages on
Word2Vec framework to embed ICD codes into vector-valued representation. We
then propose a sequential algorithm for case-control matching on this
representation space of diagnosis codes. The novel practice of applying the
sequential matching on the vector representation lifted the matching accuracy
measured through multiple clinical outcomes. We reported the results on a
large-scale dataset to demonstrate the effectiveness of our method. For such a
large dataset where most clinical information has been codified, the new method
is particularly relevant.
|
Combinatorial semi-bandit with known covariance | The combinatorial stochastic semi-bandit problem is an extension of the
classical multi-armed bandit problem in which an algorithm pulls more than one
arm at each stage and the rewards of all pulled arms are revealed. One
difference with the single arm variant is that the dependency structure of the
arms is crucial. Previous works on this setting either used a worst-case
approach or imposed independence of the arms. We introduce a way to quantify
the dependency structure of the problem and design an algorithm that adapts to
it. The algorithm is based on linear regression and the analysis develops
techniques from the linear bandit literature. By comparing its performance to a
new lower bound, we prove that it is optimal, up to a poly-logarithmic factor
in the number of pulled arms.
|
Microseismic events enhancement and detection in sensor arrays using
autocorrelation based filtering | Passive microseismic data are commonly buried in noise, which presents a
significant challenge for signal detection and recovery. For recordings from a
surface sensor array where each trace contains a time-delayed arrival from the
event, we propose an autocorrelation-based stacking method that designs a
denoising filter from all the traces, as well as a multi-channel detection
scheme. This approach circumvents the issue of time aligning the traces prior
to stacking because every trace's autocorrelation is centered at zero in the
lag domain. The effect of white noise is concentrated near zero lag, so the
filter design requires a predictable adjustment of the zero-lag value.
Truncation of the autocorrelation is employed to smooth the impulse response of
the denoising filter. In order to extend the applicability of the algorithm, we
also propose a noise prewhitening scheme that addresses cases with colored
noise. The simplicity and robustness of this method are validated with
synthetic and real seismic traces.
|
Invariant Representations for Noisy Speech Recognition | Modern automatic speech recognition (ASR) systems need to be robust under
acoustic variability arising from environmental, speaker, channel, and
recording conditions. Ensuring such robustness to variability is a challenge in
modern day neural network-based ASR systems, especially when all types of
variability are not seen during training. We attempt to address this problem by
encouraging the neural network acoustic model to learn invariant feature
representations. We use ideas from recent research on image generation using
Generative Adversarial Networks and domain adaptation ideas extending
adversarial gradient-based training. A recent work from Ganin et al. proposes
to use adversarial training for image domain adaptation by using an
intermediate representation from the main target classification network to
deteriorate the domain classifier performance through a separate neural
network. Our work focuses on investigating neural architectures which produce
representations invariant to noise conditions for ASR. We evaluate the proposed
architecture on the Aurora-4 task, a popular benchmark for noise robust ASR. We
show that our method generalizes better than the standard multi-condition
training especially when only a few noise categories are seen during training.
|
A Probabilistic Framework for Deep Learning | We develop a probabilistic framework for deep learning based on the Deep
Rendering Mixture Model (DRMM), a new generative probabilistic model that
explicitly capture variations in data due to latent task nuisance variables. We
demonstrate that max-sum inference in the DRMM yields an algorithm that exactly
reproduces the operations in deep convolutional neural networks (DCNs),
providing a first principles derivation. Our framework provides new insights
into the successes and shortcomings of DCNs as well as a principled route to
their improvement. DRMM training via the Expectation-Maximization (EM)
algorithm is a powerful alternative to DCN back-propagation, and initial
training results are promising. Classification based on the DRMM and other
variants outperforms DCNs in supervised digit classification, training 2-3x
faster while achieving similar accuracy. Moreover, the DRMM is applicable to
semi-supervised and unsupervised learning tasks, achieving results that are
state-of-the-art in several categories on the MNIST benchmark and comparable to
state of the art on the CIFAR10 benchmark.
|
Semi-Supervised Learning with the Deep Rendering Mixture Model | Semi-supervised learning algorithms reduce the high cost of acquiring labeled
training data by using both labeled and unlabeled data during learning. Deep
Convolutional Networks (DCNs) have achieved great success in supervised tasks
and as such have been widely employed in the semi-supervised learning. In this
paper we leverage the recently developed Deep Rendering Mixture Model (DRMM), a
probabilistic generative model that models latent nuisance variation, and whose
inference algorithm yields DCNs. We develop an EM algorithm for the DRMM to
learn from both labeled and unlabeled data. Guided by the theory of the DRMM,
we introduce a novel non-negativity constraint and a variational inference
term. We report state-of-the-art performance on MNIST and SVHN and competitive
results on CIFAR10. We also probe deeper into how a DRMM trained in a
semi-supervised setting represents latent nuisance variation using
synthetically rendered images. Taken together, our work provides a unified
framework for supervised, unsupervised, and semi-supervised learning.
|
Segmental Convolutional Neural Networks for Detection of Cardiac
Abnormality With Noisy Heart Sound Recordings | Heart diseases constitute a global health burden, and the problem is
exacerbated by the error-prone nature of listening to and interpreting heart
sounds. This motivates the development of automated classification to screen
for abnormal heart sounds. Existing machine learning-based systems achieve
accurate classification of heart sound recordings but rely on expert features
that have not been thoroughly evaluated on noisy recordings. Here we propose a
segmental convolutional neural network architecture that achieves automatic
feature learning from noisy heart sound recordings. Our experiments show that
our best model, trained on noisy recording segments acquired with an existing
hidden semi-markov model-based approach, attains a classification accuracy of
87.5% on the 2016 PhysioNet/CinC Challenge dataset, compared to the 84.6%
accuracy of the state-of-the-art statistical classifier trained and evaluated
on the same dataset. Our results indicate the potential of using neural
network-based methods to increase the accuracy of automated classification of
heart sound recordings for improved screening of heart diseases.
|
Core Sampling Framework for Pixel Classification | The intermediate map responses of a Convolutional Neural Network (CNN)
contain information about an image that can be used to extract contextual
knowledge about it. In this paper, we present a core sampling framework that is
able to use these activation maps from several layers as features to another
neural network using transfer learning to provide an understanding of an input
image. Our framework creates a representation that combines features from the
test data and the contextual knowledge gained from the responses of a
pretrained network, processes it and feeds it to a separate Deep Belief
Network. We use this representation to extract more information from an image
at the pixel level, hence gaining understanding of the whole image. We
experimentally demonstrate the usefulness of our framework using a pretrained
VGG-16 model to perform segmentation on the BAERI dataset of Synthetic Aperture
Radar(SAR) imagery and the CAMVID dataset.
|
Local Group Invariant Representations via Orbit Embeddings | Invariance to nuisance transformations is one of the desirable properties of
effective representations. We consider transformations that form a \emph{group}
and propose an approach based on kernel methods to derive local group invariant
representations. Locality is achieved by defining a suitable probability
distribution over the group which in turn induces distributions in the input
feature space. We learn a decision function over these distributions by
appealing to the powerful framework of kernel methods and generate local
invariant random feature maps via kernel approximations. We show uniform
convergence bounds for kernel approximation and provide excess risk bounds for
learning with these features. We evaluate our method on three real datasets,
including Rotated MNIST and CIFAR-10, and observe that it outperforms competing
kernel based approaches. The proposed method also outperforms deep CNN on
Rotated-MNIST and performs comparably to the recently proposed
group-equivariant CNN.
|
Statistical and Computational Guarantees of Lloyd's Algorithm and its
Variants | Clustering is a fundamental problem in statistics and machine learning.
Lloyd's algorithm, proposed in 1957, is still possibly the most widely used
clustering algorithm in practice due to its simplicity and empirical
performance. However, there has been little theoretical investigation on the
statistical and computational guarantees of Lloyd's algorithm. This paper is an
attempt to bridge this gap between practice and theory. We investigate the
performance of Lloyd's algorithm on clustering sub-Gaussian mixtures. Under an
appropriate initialization for labels or centers, we show that Lloyd's
algorithm converges to an exponentially small clustering error after an order
of $\log n$ iterations, where $n$ is the sample size. The error rate is shown
to be minimax optimal. For the two-mixture case, we only require the
initializer to be slightly better than random guess.
In addition, we extend the Lloyd's algorithm and its analysis to community
detection and crowdsourcing, two problems that have received a lot of attention
recently in statistics and machine learning. Two variants of Lloyd's algorithm
are proposed respectively for community detection and crowdsourcing. On the
theoretical side, we provide statistical and computational guarantees of the
two algorithms, and the results improve upon some previous signal-to-noise
ratio conditions in literature for both problems. Experimental results on
simulated and real data sets demonstrate competitive performance of our
algorithms to the state-of-the-art methods.
|
Predictive Business Process Monitoring with LSTM Neural Networks | Predictive business process monitoring methods exploit logs of completed
cases of a process in order to make predictions about running cases thereof.
Existing methods in this space are tailor-made for specific prediction tasks.
Moreover, their relative accuracy is highly sensitive to the dataset at hand,
thus requiring users to engage in trial-and-error and tuning when applying them
in a specific setting. This paper investigates Long Short-Term Memory (LSTM)
neural networks as an approach to build consistently accurate models for a wide
range of predictive process monitoring tasks. First, we show that LSTMs
outperform existing techniques to predict the next event of a running case and
its timestamp. Next, we show how to use models for predicting the next task in
order to predict the full continuation of a running case. Finally, we apply the
same approach to predict the remaining time, and show that this approach
outperforms existing tailor-made methods.
|
Mode Regularized Generative Adversarial Networks | Although Generative Adversarial Networks achieve state-of-the-art results on
a variety of generative tasks, they are regarded as highly unstable and prone
to miss modes. We argue that these bad behaviors of GANs are due to the very
particular functional shape of the trained discriminators in high dimensional
spaces, which can easily make training stuck or push probability mass in the
wrong direction, towards that of higher concentration than that of the data
generating distribution. We introduce several ways of regularizing the
objective, which can dramatically stabilize the training of GAN models. We also
show that our regularizers can help the fair distribution of probability mass
across the modes of the data generating distribution, during the early phases
of training and thus providing a unified solution to the missing modes problem.
|
Measuring the non-asymptotic convergence of sequential Monte Carlo
samplers using probabilistic programming | A key limitation of sampling algorithms for approximate inference is that it
is difficult to quantify their approximation error. Widely used sampling
schemes, such as sequential importance sampling with resampling and
Metropolis-Hastings, produce output samples drawn from a distribution that may
be far from the target posterior distribution. This paper shows how to
upper-bound the symmetric KL divergence between the output distribution of a
broad class of sequential Monte Carlo (SMC) samplers and their target posterior
distributions, subject to assumptions about the accuracy of a separate
gold-standard sampler. The proposed method applies to samplers that combine
multiple particles, multinomial resampling, and rejuvenation kernels. The
experiments show the technique being used to estimate bounds on the divergence
of SMC samplers for posterior inference in a Bayesian linear regression model
and a Dirichlet process mixture model.
|
Model-based Adversarial Imitation Learning | Generative adversarial learning is a popular new approach to training
generative models which has been proven successful for other related problems
as well. The general idea is to maintain an oracle $D$ that discriminates
between the expert's data distribution and that of the generative model $G$.
The generative model is trained to capture the expert's distribution by
maximizing the probability of $D$ misclassifying the data it generates.
Overall, the system is \emph{differentiable} end-to-end and is trained using
basic backpropagation. This type of learning was successfully applied to the
problem of policy imitation in a model-free setup. However, a model-free
approach does not allow the system to be differentiable, which requires the use
of high-variance gradient estimations. In this paper we introduce the Model
based Adversarial Imitation Learning (MAIL) algorithm. A model-based approach
for the problem of adversarial imitation learning. We show how to use a forward
model to make the system fully differentiable, which enables us to train
policies using the (stochastic) gradient of $D$. Moreover, our approach
requires relatively few environment interactions, and fewer hyper-parameters to
tune. We test our method on the MuJoCo physics simulator and report initial
results that surpass the current state-of-the-art.
|
Fast Adaptation in Generative Models with Generative Matching Networks | Despite recent advances, the remaining bottlenecks in deep generative models
are necessity of extensive training and difficulties with generalization from
small number of training examples. We develop a new generative model called
Generative Matching Network which is inspired by the recently proposed matching
networks for one-shot learning in discriminative tasks. By conditioning on the
additional input dataset, our model can instantly learn new concepts that were
not available in the training data but conform to a similar generative process.
The proposed framework does not explicitly restrict diversity of the
conditioning data and also does not require an extensive inference procedure
for training or adaptation. Our experiments on the Omniglot dataset demonstrate
that Generative Matching Networks significantly improve predictive performance
on the fly as more additional data is available and outperform existing state
of the art conditional generative models.
|
A Communication-Efficient Parallel Method for Group-Lasso | Group-Lasso (gLasso) identifies important explanatory factors in predicting
the response variable by considering the grouping structure over input
variables. However, most existing algorithms for gLasso are not scalable to
deal with large-scale datasets, which are becoming a norm in many applications.
In this paper, we present a divide-and-conquer based parallel algorithm
(DC-gLasso) to scale up gLasso in the tasks of regression with grouping
structures. DC-gLasso only needs two iterations to collect and aggregate the
local estimates on subsets of the data, and is provably correct to recover the
true model under certain conditions. We further extend it to deal with
overlappings between groups. Empirical results on a wide range of synthetic and
real-world datasets show that DC-gLasso can significantly improve the time
efficiency without sacrificing regression accuracy.
|
Large-Margin Softmax Loss for Convolutional Neural Networks | Cross-entropy loss together with softmax is arguably one of the most common
used supervision components in convolutional neural networks (CNNs). Despite
its simplicity, popularity and excellent performance, the component does not
explicitly encourage discriminative learning of features. In this paper, we
propose a generalized large-margin softmax (L-Softmax) loss which explicitly
encourages intra-class compactness and inter-class separability between learned
features. Moreover, L-Softmax not only can adjust the desired margin but also
can avoid overfitting. We also show that the L-Softmax loss can be optimized by
typical stochastic gradient descent. Extensive experiments on four benchmark
datasets demonstrate that the deeply-learned features with L-softmax loss
become more discriminative, hence significantly boosting the performance on a
variety of visual classification and verification tasks.
|
Spatially Adaptive Computation Time for Residual Networks | This paper proposes a deep learning architecture based on Residual Network
that dynamically adjusts the number of executed layers for the regions of the
image. This architecture is end-to-end trainable, deterministic and
problem-agnostic. It is therefore applicable without any modifications to a
wide range of computer vision problems such as image classification, object
detection and image segmentation. We present experimental results showing that
this model improves the computational efficiency of Residual Networks on the
challenging ImageNet classification and COCO object detection datasets.
Additionally, we evaluate the computation time maps on the visual saliency
dataset cat2000 and find that they correlate surprisingly well with human eye
fixation positions.
|
Extend natural neighbor: a novel classification method with
self-adaptive neighborhood parameters in different stages | Various kinds of k-nearest neighbor (KNN) based classification methods are
the bases of many well-established and high-performance pattern-recognition
techniques, but both of them are vulnerable to their parameter choice.
Essentially, the challenge is to detect the neighborhood of various data sets,
while utterly ignorant of the data characteristic. This article introduces a
new supervised classification method: the extend natural neighbor (ENaN)
method, and shows that it provides a better classification result without
choosing the neighborhood parameter artificially. Unlike the original KNN based
method which needs a prior k, the ENaNE method predicts different k in
different stages. Therefore, the ENaNE method is able to learn more from
flexible neighbor information both in training stage and testing stage, and
provide a better classification result.
|
Robust Low-Complexity Randomized Methods for Locating Outliers in Large
Matrices | This paper examines the problem of locating outlier columns in a large,
otherwise low-rank matrix, in settings where {}{the data} are noisy, or where
the overall matrix has missing elements. We propose a randomized two-step
inference framework, and establish sufficient conditions on the required sample
complexities under which these methods succeed (with high probability) in
accurately locating the outliers for each task. Comprehensive numerical
experimental results are provided to verify the theoretical bounds and
demonstrate the computational efficiency of the proposed algorithm.
|
An Information-theoretic Approach to Machine-oriented Music
Summarization | Music summarization allows for higher efficiency in processing, storage, and
sharing of datasets. Machine-oriented approaches, being agnostic to human
consumption, optimize these aspects even further. Such summaries have already
been successfully validated in some MIR tasks. We now generalize previous
conclusions by evaluating the impact of generic summarization of music from a
probabilistic perspective. We estimate Gaussian distributions for original and
summarized songs and compute their relative entropy, in order to measure
information loss incurred by summarization. Our results suggest that relative
entropy is a good predictor of summarization performance in the context of
tasks relying on a bag-of-features model. Based on this observation, we further
propose a straightforward yet expressive summarizer, which minimizes relative
entropy with respect to the original song, that objectively outperforms
previous methods and is better suited to avoid potential copyright issues.
|
Improving the Performance of Neural Machine Translation Involving
Morphologically Rich Languages | The advent of the attention mechanism in neural machine translation models
has improved the performance of machine translation systems by enabling
selective lookup into the source sentence. In this paper, the efficiencies of
translation using bidirectional encoder attention decoder models were studied
with respect to translation involving morphologically rich languages. The
English - Tamil language pair was selected for this analysis. First, the use of
Word2Vec embedding for both the English and Tamil words improved the
translation results by 0.73 BLEU points over the baseline RNNSearch model with
4.84 BLEU score. The use of morphological segmentation before word
vectorization to split the morphologically rich Tamil words into their
respective morphemes before the translation, caused a reduction in the target
vocabulary size by a factor of 8. Also, this model (RNNMorph) improved the
performance of neural machine translation by 7.05 BLEU points over the
RNNSearch model used over the same corpus. Since the BLEU evaluation of the
RNNMorph model might be unreliable due to an increase in the number of matching
tokens per sentence, the performances of the translations were also compared by
means of human evaluation metrics of adequacy, fluency and relative ranking.
Further, the use of morphological segmentation also improved the efficacy of
the attention mechanism.
|
Interactive Elicitation of Knowledge on Feature Relevance Improves
Predictions in Small Data Sets | Providing accurate predictions is challenging for machine learning algorithms
when the number of features is larger than the number of samples in the data.
Prior knowledge can improve machine learning models by indicating relevant
variables and parameter values. Yet, this prior knowledge is often tacit and
only available from domain experts. We present a novel approach that uses
interactive visualization to elicit the tacit prior knowledge and uses it to
improve the accuracy of prediction models. The main component of our approach
is a user model that models the domain expert's knowledge of the relevance of
different features for a prediction task. In particular, based on the expert's
earlier input, the user model guides the selection of the features on which to
elicit user's knowledge next. The results of a controlled user study show that
the user model significantly improves prior knowledge elicitation and
prediction accuracy, when predicting the relative citation counts of scientific
documents in a specific domain.
|
Bridging Medical Data Inference to Achilles Tendon Rupture
Rehabilitation | Imputing incomplete medical tests and predicting patient outcomes are crucial
for guiding the decision making for therapy, such as after an Achilles Tendon
Rupture (ATR). We formulate the problem of data imputation and prediction for
ATR relevant medical measurements into a recommender system framework. By
applying MatchBox, which is a collaborative filtering approach, on a real
dataset collected from 374 ATR patients, we aim at offering personalized
medical data imputation and prediction. In this work, we show the feasibility
of this approach and discuss potential research directions by conducting
initial qualitative evaluations.
|
Prediction with a Short Memory | We consider the problem of predicting the next observation given a sequence
of past observations, and consider the extent to which accurate prediction
requires complex algorithms that explicitly leverage long-range dependencies.
Perhaps surprisingly, our positive results show that for a broad class of
sequences, there is an algorithm that predicts well on average, and bases its
predictions only on the most recent few observation together with a set of
simple summary statistics of the past observations. Specifically, we show that
for any distribution over observations, if the mutual information between past
observations and future observations is upper bounded by $I$, then a simple
Markov model over the most recent $I/\epsilon$ observations obtains expected KL
error $\epsilon$---and hence $\ell_1$ error $\sqrt{\epsilon}$---with respect to
the optimal predictor that has access to the entire past and knows the data
generating distribution. For a Hidden Markov Model with $n$ hidden states, $I$
is bounded by $\log n$, a quantity that does not depend on the mixing time, and
we show that the trivial prediction algorithm based on the empirical
frequencies of length $O(\log n/\epsilon)$ windows of observations achieves
this error, provided the length of the sequence is $d^{\Omega(\log
n/\epsilon)}$, where $d$ is the size of the observation alphabet.
We also establish that this result cannot be improved upon, even for the
class of HMMs, in the following two senses: First, for HMMs with $n$ hidden
states, a window length of $\log n/\epsilon$ is information-theoretically
necessary to achieve expected $\ell_1$ error $\sqrt{\epsilon}$. Second, the
$d^{\Theta(\log n/\epsilon)}$ samples required to estimate the Markov model for
an observation alphabet of size $d$ is necessary for any computationally
tractable learning algorithm, assuming the hardness of strongly refuting a
certain class of CSPs.
|
Predicting brain age with deep learning from raw imaging data results in
a reliable and heritable biomarker | Machine learning analysis of neuroimaging data can accurately predict
chronological age in healthy people and deviations from healthy brain ageing
have been associated with cognitive impairment and disease. Here we sought to
further establish the credentials of "brain-predicted age" as a biomarker of
individual differences in the brain ageing process, using a predictive
modelling approach based on deep learning, and specifically convolutional
neural networks (CNN), and applied to both pre-processed and raw T1-weighted
MRI data. Firstly, we aimed to demonstrate the accuracy of CNN brain-predicted
age using a large dataset of healthy adults (N = 2001). Next, we sought to
establish the heritability of brain-predicted age using a sample of monozygotic
and dizygotic female twins (N = 62). Thirdly, we examined the test-retest and
multi-centre reliability of brain-predicted age using two samples
(within-scanner N = 20; between-scanner N = 11). CNN brain-predicted ages were
generated and compared to a Gaussian Process Regression (GPR) approach, on all
datasets. Input data were grey matter (GM) or white matter (WM) volumetric maps
generated by Statistical Parametric Mapping (SPM) or raw data. Brain-predicted
age represents an accurate, highly reliable and genetically-valid phenotype,
that has potential to be used as a biomarker of brain ageing. Moreover, age
predictions can be accurately generated on raw T1-MRI data, substantially
reducing computation time for novel data, bringing the process closer to giving
real-time information on brain health in clinical settings.
|
Towards Information-Seeking Agents | We develop a general problem setting for training and testing the ability of
agents to gather information efficiently. Specifically, we present a collection
of tasks in which success requires searching through a partially-observed
environment, for fragments of information which can be pieced together to
accomplish various goals. We combine deep architectures with techniques from
reinforcement learning to develop agents that solve our tasks. We shape the
behavior of these agents by combining extrinsic and intrinsic rewards. We
empirically demonstrate that these agents learn to search actively and
intelligently for new information to reduce their uncertainty, and to exploit
information they have already acquired.
|
Evaluating the Performance of ANN Prediction System at Shanghai Stock
Market in the Period 21-Sep-2016 to 11-Oct-2016 | This research evaluates the performance of an Artificial Neural Network based
prediction system that was employed on the Shanghai Stock Exchange for the
period 21-Sep-2016 to 11-Oct-2016. It is a follow-up to a previous paper in
which the prices were predicted and published before September 21. Stock market
price prediction remains an important quest for investors and researchers. This
research used an Artificial Intelligence system, being an Artificial Neural
Network that is feedforward multi-layer perceptron with error backpropagation
for prediction, unlike other methods such as technical, fundamental or time
series analysis. While these alternative methods tend to guide on trends and
not the exact likely prices, neural networks on the other hand have the ability
to predict the real value prices, as was done on this research. Nonetheless,
determination of suitable network parameters remains a challenge in neural
network design, with this research settling on a configuration of 5:21:21:1
with 80% training data or 4-year of training data as a good enough model for
stock prediction, as already determined in a previous research by the author.
The comparative results indicate that neural network can predict typical stock
market prices with mean absolute percentage errors that are as low as 1.95%
over the ten prediction instances that was studied in this research.
|
Towards better decoding and language model integration in sequence to
sequence models | The recently proposed Sequence-to-Sequence (seq2seq) framework advocates
replacing complex data processing pipelines, such as an entire automatic speech
recognition system, with a single neural network trained in an end-to-end
fashion. In this contribution, we analyse an attention-based seq2seq speech
recognition system that directly transcribes recordings into characters. We
observe two shortcomings: overconfidence in its predictions and a tendency to
produce incomplete transcriptions when language models are used. We propose
practical solutions to both problems achieving competitive speaker independent
word error rates on the Wall Street Journal dataset: without separate language
models we reach 10.6% WER, while together with a trigram language model, we
reach 6.7% WER.
|
A note on the triangle inequality for the Jaccard distance | Two simple proofs of the triangle inequality for the Jaccard distance in
terms of nonnegative, monotone, submodular functions are given and discussed.
|
CrowdMI: Multiple Imputation via Crowdsourcing | Can humans impute missing data with similar proficiency as machines? This is
the question we aim to answer in this paper. We present a novel idea of
converting observations with missing data in to a survey questionnaire, which
is presented to crowdworkers for completion. We replicate a multiple imputation
framework by having multiple unique crowdworkers complete our questionnaire.
Experimental results demonstrate that using our method, it is possible to
generate valid imputations for qualitative and quantitative missing data, with
results comparable to imputations generated by complex statistical models.
|
Scalable Influence Maximization for Multiple Products in Continuous-Time
Diffusion Networks | A typical viral marketing model identifies influential users in a social
network to maximize a single product adoption assuming unlimited user
attention, campaign budgets, and time. In reality, multiple products need
campaigns, users have limited attention, convincing users incurs costs, and
advertisers have limited budgets and expect the adoptions to be maximized soon.
Facing these user, monetary, and timing constraints, we formulate the problem
as a submodular maximization task in a continuous-time diffusion model under
the intersection of a matroid and multiple knapsack constraints. We propose a
randomized algorithm estimating the user influence in a network
($|\mathcal{V}|$ nodes, $|\mathcal{E}|$ edges) to an accuracy of $\epsilon$
with $n=\mathcal{O}(1/\epsilon^2)$ randomizations and
$\tilde{\mathcal{O}}(n|\mathcal{E}|+n|\mathcal{V}|)$ computations. By
exploiting the influence estimation algorithm as a subroutine, we develop an
adaptive threshold greedy algorithm achieving an approximation factor $k_a/(2+2
k)$ of the optimal when $k_a$ out of the $k$ knapsack constraints are active.
Extensive experiments on networks of millions of nodes demonstrate that the
proposed algorithms achieve the state-of-the-art in terms of effectiveness and
scalability.
|
Learning in the Machine: Random Backpropagation and the Deep Learning
Channel | Random backpropagation (RBP) is a variant of the backpropagation algorithm
for training neural networks, where the transpose of the forward matrices are
replaced by fixed random matrices in the calculation of the weight updates. It
is remarkable both because of its effectiveness, in spite of using random
matrices to communicate error information, and because it completely removes
the taxing requirement of maintaining symmetric weights in a physical neural
system. To better understand random backpropagation, we first connect it to the
notions of local learning and learning channels. Through this connection, we
derive several alternatives to RBP, including skipped RBP (SRPB), adaptive RBP
(ARBP), sparse RBP, and their combinations (e.g. ASRBP) and analyze their
computational complexity. We then study their behavior through simulations
using the MNIST and CIFAR-10 bechnmark datasets. These simulations show that
most of these variants work robustly, almost as well as backpropagation, and
that multiplication by the derivatives of the activation functions is
important. As a follow-up, we study also the low-end of the number of bits
required to communicate error information over the learning channel. We then
provide partial intuitive explanations for some of the remarkable properties of
RBP and its variations. Finally, we prove several mathematical results,
including the convergence to fixed points of linear chains of arbitrary length,
the convergence to fixed points of linear autoencoders with decorrelated data,
the long-term existence of solutions for linear systems with a single hidden
layer and convergence in special cases, and the convergence to fixed points of
non-linear chains, when the derivative of the activation functions is included.
|
Controlling Robot Morphology from Incomplete Measurements | Mobile robots with complex morphology are essential for traversing rough
terrains in Urban Search & Rescue missions (USAR). Since teleoperation of the
complex morphology causes high cognitive load of the operator, the morphology
is controlled autonomously. The autonomous control measures the robot state and
surrounding terrain which is usually only partially observable, and thus the
data are often incomplete. We marginalize the control over the missing
measurements and evaluate an explicit safety condition. If the safety condition
is violated, tactile terrain exploration by the body-mounted robotic arm
gathers the missing data.
|
Coupling Distributed and Symbolic Execution for Natural Language Queries | Building neural networks to query a knowledge base (a table) with natural
language is an emerging research topic in deep learning. An executor for table
querying typically requires multiple steps of execution because queries may
have complicated structures. In previous studies, researchers have developed
either fully distributed executors or symbolic executors for table querying. A
distributed executor can be trained in an end-to-end fashion, but is weak in
terms of execution efficiency and explicit interpretability. A symbolic
executor is efficient in execution, but is very difficult to train especially
at initial stages. In this paper, we propose to couple distributed and symbolic
execution for natural language queries, where the symbolic executor is
pretrained with the distributed executor's intermediate execution results in a
step-by-step fashion. Experiments show that our approach significantly
outperforms both distributed and symbolic executors, exhibiting high accuracy,
high learning efficiency, high execution efficiency, and high interpretability.
|
Protein-Ligand Scoring with Convolutional Neural Networks | Computational approaches to drug discovery can reduce the time and cost
associated with experimental assays and enable the screening of novel
chemotypes. Structure-based drug design methods rely on scoring functions to
rank and predict binding affinities and poses. The ever-expanding amount of
protein-ligand binding and structural data enables the use of deep machine
learning techniques for protein-ligand scoring.
We describe convolutional neural network (CNN) scoring functions that take as
input a comprehensive 3D representation of a protein-ligand interaction. A CNN
scoring function automatically learns the key features of protein-ligand
interactions that correlate with binding. We train and optimize our CNN scoring
functions to discriminate between correct and incorrect binding poses and known
binders and non-binders. We find that our CNN scoring function outperforms the
AutoDock Vina scoring function when ranking poses both for pose prediction and
virtual screening.
|
Improved generator objectives for GANs | We present a framework to understand GAN training as alternating density
ratio estimation and approximate divergence minimization. This provides an
interpretation for the mismatched GAN generator and discriminator objectives
often used in practice, and explains the problem of poor sample diversity. We
also derive a family of generator objectives that target arbitrary
$f$-divergences without minimizing a lower bound, and use them to train
generative image models that target either improved sample quality or greater
sample diversity.
|
Interactive Prior Elicitation of Feature Similarities for Small Sample
Size Prediction | Regression under the "small $n$, large $p$" conditions, of small sample size
$n$ and large number of features $p$ in the learning data set, is a recurring
setting in which learning from data is difficult. With prior knowledge about
relationships of the features, $p$ can effectively be reduced, but explicating
such prior knowledge is difficult for experts. In this paper we introduce a new
method for eliciting expert prior knowledge about the similarity of the roles
of features in the prediction task. The key idea is to use an interactive
multidimensional-scaling (MDS) type scatterplot display of the features to
elicit the similarity relationships, and then use the elicited relationships in
the prior distribution of prediction parameters. Specifically, for learning to
predict a target variable with Bayesian linear regression, the feature
relationships are used to construct a Gaussian prior with a full covariance
matrix for the regression coefficients. Evaluation of our method in experiments
with simulated and real users on text data confirm that prior elicitation of
feature similarities improves prediction accuracy. Furthermore, elicitation
with an interactive scatterplot display outperforms straightforward elicitation
where the users choose feature pairs from a feature list.
|
The Physical Systems Behind Optimization Algorithms | We use differential equations based approaches to provide some {\it
\textbf{physics}} insights into analyzing the dynamics of popular optimization
algorithms in machine learning. In particular, we study gradient descent,
proximal gradient descent, coordinate gradient descent, proximal coordinate
gradient, and Newton's methods as well as their Nesterov's accelerated variants
in a unified framework motivated by a natural connection of optimization
algorithms to physical systems. Our analysis is applicable to more general
algorithms and optimization problems {\it \textbf{beyond}} convexity and strong
convexity, e.g. Polyak-\L ojasiewicz and error bound conditions (possibly
nonconvex).
|
Task-Guided and Path-Augmented Heterogeneous Network Embedding for
Author Identification | In this paper, we study the problem of author identification under
double-blind review setting, which is to identify potential authors given
information of an anonymized paper. Different from existing approaches that
rely heavily on feature engineering, we propose to use network embedding
approach to address the problem, which can automatically represent nodes into
lower dimensional feature vectors. However, there are two major limitations in
recent studies on network embedding: (1) they are usually general-purpose
embedding methods, which are independent of the specific tasks; and (2) most of
these approaches can only deal with homogeneous networks, where the
heterogeneity of the network is ignored. Hence, challenges faced here are two
folds: (1) how to embed the network under the guidance of the author
identification task, and (2) how to select the best type of information due to
the heterogeneity of the network.
To address the challenges, we propose a task-guided and path-augmented
heterogeneous network embedding model. In our model, nodes are first embedded
as vectors in latent feature space. Embeddings are then shared and jointly
trained according to task-specific and network-general objectives. We extend
the existing unsupervised network embedding to incorporate meta paths in
heterogeneous networks, and select paths according to the specific task. The
guidance from author identification task for network embedding is provided both
explicitly in joint training and implicitly during meta path selection. Our
experiments demonstrate that by using path-augmented network embedding with
task guidance, our model can obtain significantly better accuracy at
identifying the true authors comparing to existing methods.
|
Tensor-Dictionary Learning with Deep Kruskal-Factor Analysis | A multi-way factor analysis model is introduced for tensor-variate data of
any order. Each data item is represented as a (sparse) sum of Kruskal
decompositions, a Kruskal-factor analysis (KFA). KFA is nonparametric and can
infer both the tensor-rank of each dictionary atom and the number of dictionary
atoms. The model is adapted for online learning, which allows dictionary
learning on large data sets. After KFA is introduced, the model is extended to
a deep convolutional tensor-factor analysis, supervised by a Bayesian SVM. The
experiments section demonstrates the improvement of KFA over vectorized
approaches (e.g., BPFA), tensor decompositions, and convolutional neural
networks (CNN) in multi-way denoising, blind inpainting, and image
classification. The improvement in PSNR for the inpainting results over other
methods exceeds 1dB in several cases and we achieve state of the art results on
Caltech101 image classification.
|
Learning Representations by Stochastic Meta-Gradient Descent in Neural
Networks | Representations are fundamental to artificial intelligence. The performance
of a learning system depends on the type of representation used for
representing the data. Typically, these representations are hand-engineered
using domain knowledge. More recently, the trend is to learn these
representations through stochastic gradient descent in multi-layer neural
networks, which is called backprop. Learning the representations directly from
the incoming data stream reduces the human labour involved in designing a
learning system. More importantly, this allows in scaling of a learning system
for difficult tasks. In this paper, we introduce a new incremental learning
algorithm called crossprop, which learns incoming weights of hidden units based
on the meta-gradient descent approach, that was previously introduced by Sutton
(1992) and Schraudolph (1999) for learning step-sizes. The final update
equation introduces an additional memory parameter for each of these weights
and generalizes the backprop update equation. From our experiments, we show
that crossprop learns and reuses its feature representation while tackling new
and unseen tasks whereas backprop relearns a new feature representation.
|
A Review of Intelligent Practices for Irrigation Prediction | Population growth and increasing droughts are creating unprecedented strain
on the continued availability of water resources. Since irrigation is a major
consumer of fresh water, wastage of resources in this sector could have strong
consequences. To address this issue, irrigation water management and prediction
techniques need to be employed effectively and should be able to account for
the variabilities present in the environment. The different techniques surveyed
in this paper can be classified into two categories: computational and
statistical. Computational methods deal with scientific correlations between
physical parameters whereas statistical methods involve specific prediction
algorithms that can be used to automate the process of irrigation water
prediction. These algorithms interpret semantic relationships between the
various parameters of temperature, pressure, evapotranspiration etc. and store
them as numerical precomputed entities specific to the conditions and the area
used as the data for the training corpus used to train it. We focus on
reviewing the computational methods used to determine Evapotranspiration and
its implications. We compare the efficiencies of different data mining and
machine learning methods implemented in this area, such as Logistic Regression,
Decision Tress Classifier, SysFor, Support Vector Machine(SVM), Fuzzy Logic
techniques, Artifical Neural Networks(ANNs) and various hybrids of Genetic
Algorithms (GA) applied to irrigation prediction. We also recommend a possible
technique for the same based on its superior results in other such time series
analysis tasks.
|
Environmental Modeling Framework using Stacked Gaussian Processes | A network of independently trained Gaussian processes (StackedGP) is
introduced to obtain predictions of quantities of interest with quantified
uncertainties. The main applications of the StackedGP framework are to
integrate different datasets through model composition, enhance predictions of
quantities of interest through a cascade of intermediate predictions, and to
propagate uncertainties through emulated dynamical systems driven by uncertain
forcing variables. By using analytical first and second-order moments of a
Gaussian process with uncertain inputs using squared exponential and polynomial
kernels, approximated expectations of quantities of interests that require an
arbitrary composition of functions can be obtained. The StackedGP model is
extended to any number of layers and nodes per layer, and it provides
flexibility in kernel selection for the input nodes. The proposed nonparametric
stacked model is validated using synthetic datasets, and its performance in
model composition and cascading predictions is measured in two applications
using real data.
|
A series of maximum entropy upper bounds of the differential entropy | We present a series of closed-form maximum entropy upper bounds for the
differential entropy of a continuous univariate random variable and study the
properties of that series. We then show how to use those generic bounds for
upper bounding the differential entropy of Gaussian mixture models. This
requires to calculate the raw moments and raw absolute moments of Gaussian
mixtures in closed-form that may also be handy in statistical machine learning
and information theory. We report on our experiments and discuss on the
tightness of those bounds.
|
BaTFLED: Bayesian Tensor Factorization Linked to External Data | The vast majority of current machine learning algorithms are designed to
predict single responses or a vector of responses, yet many types of response
are more naturally organized as matrices or higher-order tensor objects where
characteristics are shared across modes. We present a new machine learning
algorithm BaTFLED (Bayesian Tensor Factorization Linked to External Data) that
predicts values in a three-dimensional response tensor using input features for
each of the dimensions. BaTFLED uses a probabilistic Bayesian framework to
learn projection matrices mapping input features for each mode into latent
representations that multiply to form the response tensor. By utilizing a
Tucker decomposition, the model can capture weights for interactions between
latent factors for each mode in a small core tensor. Priors that encourage
sparsity in the projection matrices and core tensor allow for feature selection
and model regularization. This method is shown to far outperform elastic net
and neural net models on 'cold start' tasks from data simulated in a three-mode
structure. Additionally, we apply the model to predict dose-response curves in
a panel of breast cancer cell lines treated with drug compounds that was used
as a Dialogue for Reverse Engineering Assessments and Methods (DREAM)
challenge.
|
Clipper: A Low-Latency Online Prediction Serving System | Machine learning is being deployed in a growing number of applications which
demand real-time, accurate, and robust predictions under heavy query load.
However, most machine learning frameworks and systems only address model
training and not deployment.
In this paper, we introduce Clipper, a general-purpose low-latency prediction
serving system. Interposing between end-user applications and a wide range of
machine learning frameworks, Clipper introduces a modular architecture to
simplify model deployment across frameworks and applications. Furthermore, by
introducing caching, batching, and adaptive model selection techniques, Clipper
reduces prediction latency and improves prediction throughput, accuracy, and
robustness without modifying the underlying machine learning frameworks. We
evaluate Clipper on four common machine learning benchmark datasets and
demonstrate its ability to meet the latency, accuracy, and throughput demands
of online serving applications. Finally, we compare Clipper to the TensorFlow
Serving system and demonstrate that we are able to achieve comparable
throughput and latency while enabling model composition and online learning to
improve accuracy and render more robust predictions.
|
Advancing Bayesian Optimization: The Mixed-Global-Local (MGL) Kernel and
Length-Scale Cool Down | Bayesian Optimization (BO) has become a core method for solving expensive
black-box optimization problems. While much research focussed on the choice of
the acquisition function, we focus on online length-scale adaption and the
choice of kernel function. Instead of choosing hyperparameters in view of
maximum likelihood on past data, we propose to use the acquisition function to
decide on hyperparameter adaptation more robustly and in view of the future
optimization progress. Further, we propose a particular kernel function that
includes non-stationarity and local anisotropy and thereby implicitly
integrates the efficiency of local convex optimization with global Bayesian
optimization. Comparisons to state-of-the art BO methods underline the
efficiency of these mechanisms on global optimization benchmarks.
|
Phase transitions in Restricted Boltzmann Machines with generic priors | We study Generalised Restricted Boltzmann Machines with generic priors for
units and weights, interpolating between Boolean and Gaussian variables. We
present a complete analysis of the replica symmetric phase diagram of these
systems, which can be regarded as Generalised Hopfield models. We underline the
role of the retrieval phase for both inference and learning processes and we
show that retrieval is robust for a large class of weight and unit priors,
beyond the standard Hopfield scenario. Furthermore we show how the paramagnetic
phase boundary is directly related to the optimal size of the training set
necessary for good generalisation in a teacher-student scenario of unsupervised
learning.
|
Testing Ising Models | Given samples from an unknown multivariate distribution $p$, is it possible
to distinguish whether $p$ is the product of its marginals versus $p$ being far
from every product distribution? Similarly, is it possible to distinguish
whether $p$ equals a given distribution $q$ versus $p$ and $q$ being far from
each other? These problems of testing independence and goodness-of-fit have
received enormous attention in statistics, information theory, and theoretical
computer science, with sample-optimal algorithms known in several interesting
regimes of parameters. Unfortunately, it has also been understood that these
problems become intractable in large dimensions, necessitating exponential
sample complexity.
Motivated by the exponential lower bounds for general distributions as well
as the ubiquity of Markov Random Fields (MRFs) in the modeling of
high-dimensional distributions, we initiate the study of distribution testing
on structured multivariate distributions, and in particular the prototypical
example of MRFs: the Ising Model. We demonstrate that, in this structured
setting, we can avoid the curse of dimensionality, obtaining sample and time
efficient testers for independence and goodness-of-fit. One of the key
technical challenges we face along the way is bounding the variance of
functions of the Ising model.
|
Optimal mean-based algorithms for trace reconstruction | In the (deletion-channel) trace reconstruction problem, there is an unknown
$n$-bit source string $x$. An algorithm is given access to independent traces
of $x$, where a trace is formed by deleting each bit of~$x$ independently with
probability~$\delta$. The goal of the algorithm is to recover~$x$ exactly (with
high probability), while minimizing samples (number of traces) and running
time.
Previously, the best known algorithm for the trace reconstruction problem was
due to Holenstein~et~al.; it uses $\exp(\tilde{O}(n^{1/2}))$ samples and
running time for any fixed $0 < \delta < 1$. It is also what we call a
"mean-based algorithm", meaning that it only uses the empirical means of the
individual bits of the traces. Holenstein~et~al.~also gave a lower bound,
showing that any mean-based algorithm must use at least $n^{\tilde{\Omega}(\log
n)}$ samples.
In this paper we improve both of these results, obtaining matching upper and
lower bounds for mean-based trace reconstruction. For any constant deletion
rate $0 < \delta < 1$, we give a mean-based algorithm that uses
$\exp(O(n^{1/3}))$ time and traces; we also prove that any mean-based algorithm
must use at least $\exp(\Omega(n^{1/3}))$ traces. In fact, we obtain matching
upper and lower bounds even for $\delta$ subconstant and $\rho := 1-\delta$
subconstant: when $(\log^3 n)/n \ll \delta \leq 1/2$ the bound is
$\exp(-\Theta(\delta n)^{1/3})$, and when $1/\sqrt{n} \ll \rho \leq 1/2$ the
bound is $\exp(-\Theta(n/\rho)^{1/3})$.
Our proofs involve estimates for the maxima of Littlewood polynomials on
complex disks. We show that these techniques can also be used to perform trace
reconstruction with random insertions and bit-flips in addition to deletions.
We also find a surprising result: for deletion probabilities $\delta > 1/2$,
the presence of insertions can actually help with trace reconstruction.
|
Testing Bayesian Networks | This work initiates a systematic investigation of testing high-dimensional
structured distributions by focusing on testing Bayesian networks -- the
prototypical family of directed graphical models. A Bayesian network is defined
by a directed acyclic graph, where we associate a random variable with each
node. The value at any particular node is conditionally independent of all the
other non-descendant nodes once its parents are fixed. Specifically, we study
the properties of identity testing and closeness testing of Bayesian networks.
Our main contribution is the first non-trivial efficient testing algorithms for
these problems and corresponding information-theoretic lower bounds. For a wide
range of parameter settings, our testing algorithms have sample complexity
sublinear in the dimension and are sample-optimal, up to constant factors.
|
Square Hellinger Subadditivity for Bayesian Networks and its
Applications to Identity Testing | We show that the square Hellinger distance between two Bayesian networks on
the same directed graph, $G$, is subadditive with respect to the neighborhoods
of $G$. Namely, if $P$ and $Q$ are the probability distributions defined by two
Bayesian networks on the same DAG, our inequality states that the square
Hellinger distance, $H^2(P,Q)$, between $P$ and $Q$ is upper bounded by the
sum, $\sum_v H^2(P_{\{v\} \cup \Pi_v}, Q_{\{v\} \cup \Pi_v})$, of the square
Hellinger distances between the marginals of $P$ and $Q$ on every node $v$ and
its parents $\Pi_v$ in the DAG. Importantly, our bound does not involve the
conditionals but the marginals of $P$ and $Q$. We derive a similar inequality
for more general Markov Random Fields.
As an application of our inequality, we show that distinguishing whether two
Bayesian networks $P$ and $Q$ on the same (but potentially unknown) DAG satisfy
$P=Q$ vs $d_{\rm TV}(P,Q)>\epsilon$ can be performed from
$\tilde{O}(|\Sigma|^{3/4(d+1)} \cdot n/\epsilon^2)$ samples, where $d$ is the
maximum in-degree of the DAG and $\Sigma$ the domain of each variable of the
Bayesian networks. If $P$ and $Q$ are defined on potentially different and
potentially unknown trees, the sample complexity becomes
$\tilde{O}(|\Sigma|^{4.5} n/\epsilon^2)$, whose dependence on $n, \epsilon$ is
optimal up to logarithmic factors. Lastly, if $P$ and $Q$ are product
distributions over $\{0,1\}^n$ and $Q$ is known, the sample complexity becomes
$O(\sqrt{n}/\epsilon^2)$, which is optimal up to constant factors.
|
Low-Rank Inducing Norms with Optimality Interpretations | Optimization problems with rank constraints appear in many diverse fields
such as control, machine learning and image analysis. Since the rank constraint
is non-convex, these problems are often approximately solved via convex
relaxations. Nuclear norm regularization is the prevailing convexifying
technique for dealing with these types of problem. This paper introduces a
family of low-rank inducing norms and regularizers which includes the nuclear
norm as a special case. A posteriori guarantees on solving an underlying rank
constrained optimization problem with these convex relaxations are provided. We
evaluate the performance of the low-rank inducing norms on three matrix
completion problems. In all examples, the nuclear norm heuristic is
outperformed by convex relaxations based on other low-rank inducing norms. For
two of the problems there exist low-rank inducing norms that succeed in
recovering the partially unknown matrix, while the nuclear norm fails. These
low-rank inducing norms are shown to be representable as semi-definite
programs. Moreover, these norms have cheaply computable proximal mappings,
which makes it possible to also solve problems of large size using first-order
methods.
|
DeepCancer: Detecting Cancer through Gene Expressions via Deep
Generative Learning | Transcriptional profiling on microarrays to obtain gene expressions has been
used to facilitate cancer diagnosis. We propose a deep generative machine
learning architecture (called DeepCancer) that learn features from unlabeled
microarray data. These models have been used in conjunction with conventional
classifiers that perform classification of the tissue samples as either being
cancerous or non-cancerous. The proposed model has been tested on two different
clinical datasets. The evaluation demonstrates that DeepCancer model achieves a
very high precision score, while significantly controlling the false positive
and false negative scores.
|
Towards deep learning with spiking neurons in energy based models with
contrastive Hebbian plasticity | In machine learning, error back-propagation in multi-layer neural networks
(deep learning) has been impressively successful in supervised and
reinforcement learning tasks. As a model for learning in the brain, however,
deep learning has long been regarded as implausible, since it relies in its
basic form on a non-local plasticity rule. To overcome this problem,
energy-based models with local contrastive Hebbian learning were proposed and
tested on a classification task with networks of rate neurons. We extended this
work by implementing and testing such a model with networks of leaky
integrate-and-fire neurons. Preliminary results indicate that it is possible to
learn a non-linear regression task with hidden layers, spiking neurons and a
local synaptic plasticity rule.
|
Optimal Generalized Decision Trees via Integer Programming | Decision trees have been a very popular class of predictive models for
decades due to their interpretability and good performance on categorical
features. However, they are not always robust and tend to overfit the data.
Additionally, if allowed to grow large, they lose interpretability. In this
paper, we present a mixed integer programming formulation to construct optimal
decision trees of a prespecified size. We take the special structure of
categorical features into account and allow combinatorial decisions (based on
subsets of values of features) at each node. Our approach can also handle
numerical features via thresholding. We show that very good accuracy can be
achieved with small trees using moderately-sized training sets. The
optimization problems we solve are tractable with modern solvers.
|
Active Learning for Speech Recognition: the Power of Gradients | In training speech recognition systems, labeling audio clips can be
expensive, and not all data is equally valuable. Active learning aims to label
only the most informative samples to reduce cost. For speech recognition,
confidence scores and other likelihood-based active learning methods have been
shown to be effective. Gradient-based active learning methods, however, are
still not well-understood. This work investigates the Expected Gradient Length
(EGL) approach in active learning for end-to-end speech recognition. We justify
EGL from a variance reduction perspective, and observe that EGL's measure of
informativeness picks novel samples uncorrelated with confidence scores.
Experimentally, we show that EGL can reduce word errors by 11\%, or
alternatively, reduce the number of samples to label by 50\%, when compared to
random sampling.
|
Generalized Deep Image to Image Regression | We present a Deep Convolutional Neural Network architecture which serves as a
generic image-to-image regressor that can be trained end-to-end without any
further machinery. Our proposed architecture: the Recursively Branched
Deconvolutional Network (RBDN) develops a cheap multi-context image
representation very early on using an efficient recursive branching scheme with
extensive parameter sharing and learnable upsampling. This multi-context
representation is subjected to a highly non-linear locality preserving
transformation by the remainder of our network comprising of a series of
convolutions/deconvolutions without any spatial downsampling. The RBDN
architecture is fully convolutional and can handle variable sized images during
inference. We provide qualitative/quantitative results on $3$ diverse tasks:
relighting, denoising and colorization and show that our proposed RBDN
architecture obtains comparable results to the state-of-the-art on each of
these tasks when used off-the-shelf without any post processing or
task-specific architectural modifications.
|
Gradient Coding | We propose a novel coding theoretic framework for mitigating stragglers in
distributed learning. We show how carefully replicating data blocks and coding
across gradients can provide tolerance to failures and stragglers for
Synchronous Gradient Descent. We implement our schemes in python (using MPI) to
run on Amazon EC2, and show how we compare against baseline approaches in
running time and generalization error.
|
Knowledge Elicitation via Sequential Probabilistic Inference for
High-Dimensional Prediction | Prediction in a small-sized sample with a large number of covariates, the
"small n, large p" problem, is challenging. This setting is encountered in
multiple applications, such as precision medicine, where obtaining additional
samples can be extremely costly or even impossible, and extensive research
effort has recently been dedicated to finding principled solutions for accurate
prediction. However, a valuable source of additional information, domain
experts, has not yet been efficiently exploited. We formulate knowledge
elicitation generally as a probabilistic inference process, where expert
knowledge is sequentially queried to improve predictions. In the specific case
of sparse linear regression, where we assume the expert has knowledge about the
values of the regression coefficients or about the relevance of the features,
we propose an algorithm and computational approximation for fast and efficient
interaction, which sequentially identifies the most informative features on
which to query expert knowledge. Evaluations of our method in experiments with
simulated and real users show improved prediction accuracy already with a small
effort from the expert.
|
An Empirical Study of ADMM for Nonconvex Problems | The alternating direction method of multipliers (ADMM) is a common
optimization tool for solving constrained and non-differentiable problems. We
provide an empirical study of the practical performance of ADMM on several
nonconvex applications, including l0 regularized linear regression, l0
regularized image denoising, phase retrieval, and eigenvector computation. Our
experiments suggest that ADMM performs well on a broad class of non-convex
problems. Moreover, recently proposed adaptive ADMM methods, which
automatically tune penalty parameters as the method runs, can improve algorithm
efficiency and solution quality compared to ADMM with a non-tuned penalty.
|
Non-negative Factorization of the Occurrence Tensor from Financial
Contracts | We propose an algorithm for the non-negative factorization of an occurrence
tensor built from heterogeneous networks. We use l0 norm to model sparse errors
over discrete values (occurrences), and use decomposed factors to model the
embedded groups of nodes. An efficient splitting method is developed to
optimize the nonconvex and nonsmooth objective. We study both synthetic
problems and a new dataset built from financial documents, resMBS.
|
Technical Report: A Generalized Matching Pursuit Approach for
Graph-Structured Sparsity | Sparsity-constrained optimization is an important and challenging problem
that has wide applicability in data mining, machine learning, and statistics.
In this paper, we focus on sparsity-constrained optimization in cases where the
cost function is a general nonlinear function and, in particular, the sparsity
constraint is defined by a graph-structured sparsity model. Existing methods
explore this problem in the context of sparse estimation in linear models. To
the best of our knowledge, this is the first work to present an efficient
approximation algorithm, namely, Graph-structured Matching Pursuit (Graph-Mp),
to optimize a general nonlinear function subject to graph-structured
constraints. We prove that our algorithm enjoys the strong guarantees analogous
to those designed for linear models in terms of convergence rate and
approximation accuracy. As a case study, we specialize Graph-Mp to optimize a
number of well-known graph scan statistic models for the connected subgraph
detection task, and empirical evidence demonstrates that our general algorithm
performs superior over state-of-the-art methods that are designed specifically
for the task of connected subgraph detection.
|
Non-Redundant Spectral Dimensionality Reduction | Spectral dimensionality reduction algorithms are widely used in numerous
domains, including for recognition, segmentation, tracking and visualization.
However, despite their popularity, these algorithms suffer from a major
limitation known as the "repeated Eigen-directions" phenomenon. That is, many
of the embedding coordinates they produce typically capture the same direction
along the data manifold. This leads to redundant and inefficient
representations that do not reveal the true intrinsic dimensionality of the
data. In this paper, we propose a general method for avoiding redundancy in
spectral algorithms. Our approach relies on replacing the orthogonality
constraints underlying those methods by unpredictability constraints.
Specifically, we require that each embedding coordinate be unpredictable (in
the statistical sense) from all previous ones. We prove that these constraints
necessarily prevent redundancy, and provide a simple technique to incorporate
them into existing methods. As we illustrate on challenging high-dimensional
scenarios, our approach produces significantly more informative and compact
representations, which improve visualization and classification tasks.
|
Lock-Free Optimization for Non-Convex Problems | Stochastic gradient descent~(SGD) and its variants have attracted much
attention in machine learning due to their efficiency and effectiveness for
optimization. To handle large-scale problems, researchers have recently
proposed several lock-free strategy based parallel SGD~(LF-PSGD) methods for
multi-core systems. However, existing works have only proved the convergence of
these LF-PSGD methods for convex problems. To the best of our knowledge, no
work has proved the convergence of the LF-PSGD methods for non-convex problems.
In this paper, we provide the theoretical proof about the convergence of two
representative LF-PSGD methods, Hogwild! and AsySVRG, for non-convex problems.
Empirical results also show that both Hogwild! and AsySVRG are convergent on
non-convex problems, which successfully verifies our theoretical results.
|
Noisy subspace clustering via matching pursuits | Sparsity-based subspace clustering algorithms have attracted significant
attention thanks to their excellent performance in practical applications. A
prominent example is the sparse subspace clustering (SSC) algorithm by
Elhamifar and Vidal, which performs spectral clustering based on an adjacency
matrix obtained by sparsely representing each data point in terms of all the
other data points via the Lasso. When the number of data points is large or the
dimension of the ambient space is high, the computational complexity of SSC
quickly becomes prohibitive. Dyer et al. observed that SSC-OMP obtained by
replacing the Lasso by the greedy orthogonal matching pursuit (OMP) algorithm
results in significantly lower computational complexity, while often yielding
comparable performance. The central goal of this paper is an analytical
performance characterization of SSC-OMP for noisy data. Moreover, we introduce
and analyze the SSC-MP algorithm, which employs matching pursuit (MP) in lieu
of OMP. Both SSC-OMP and SSC-MP are proven to succeed even when the subspaces
intersect and when the data points are contaminated by severe noise. The
clustering conditions we obtain for SSC-OMP and SSC-MP are similar to those for
SSC and for the thresholding-based subspace clustering (TSC) algorithm due to
Heckel and B\"olcskei. Analytical results in combination with numerical results
indicate that both SSC-OMP and SSC-MP with a data-dependent stopping criterion
automatically detect the dimensions of the subspaces underlying the data.
Moreover, experiments on synthetic and on real data show that SSC-MP compares
very favorably to SSC, SSC-OMP, TSC, and the nearest subspace neighbor
algorithm, both in terms of clustering performance and running time. In
addition, we find that, in contrast to SSC-OMP, the performance of SSC-MP is
very robust with respect to the choice of parameters in the stopping criteria.
|
Self-calibrating Neural Networks for Dimensionality Reduction | Recently, a novel family of biologically plausible online algorithms for
reducing the dimensionality of streaming data has been derived from the
similarity matching principle. In these algorithms, the number of output
dimensions can be determined adaptively by thresholding the singular values of
the input data matrix. However, setting such threshold requires knowing the
magnitude of the desired singular values in advance. Here we propose online
algorithms where the threshold is self-calibrating based on the singular values
computed from the existing observations. To derive these algorithms from the
similarity matching cost function we propose novel regularizers. As before,
these online algorithms can be implemented by Hebbian/anti-Hebbian neural
networks in which the learning rule depends on the chosen regularizer. We
demonstrate both mathematically and via simulation the effectiveness of these
online algorithms in various settings.
|
Kernel-based Reconstruction of Space-time Functions on Dynamic Graphs | Graph-based methods pervade the inference toolkits of numerous disciplines
including sociology, biology, neuroscience, physics, chemistry, and
engineering. A challenging problem encountered in this context pertains to
determining the attributes of a set of vertices given those of another subset
at possibly different time instants. Leveraging spatiotemporal dynamics can
drastically reduce the number of observed vertices, and hence the cost of
sampling. Alleviating the limited flexibility of existing approaches, the
present paper broadens the existing kernel-based graph function reconstruction
framework to accommodate time-evolving functions over possibly time-evolving
topologies. This approach inherits the versatility and generality of
kernel-based methods, for which no knowledge on distributions or second-order
statistics is required. Systematic guidelines are provided to construct two
families of space-time kernels with complementary strengths. The first
facilitates judicious control of regularization on a space-time frequency
plane, whereas the second can afford time-varying topologies. Batch and online
estimators are also put forth, and a novel kernel Kalman filter is developed to
obtain these estimates at affordable computational cost. Numerical tests with
real data sets corroborate the merits of the proposed methods relative to
competing alternatives.
|
FastText.zip: Compressing text classification models | We consider the problem of producing compact architectures for text
classification, such that the full model fits in a limited amount of memory.
After considering different solutions inspired by the hashing literature, we
propose a method built upon product quantization to store word embeddings.
While the original technique leads to a loss in accuracy, we adapt this method
to circumvent quantization artefacts. Our experiments carried out on several
benchmarks show that our approach typically requires two orders of magnitude
less memory than fastText while being only slightly inferior with respect to
accuracy. As a result, it outperforms the state of the art by a good margin in
terms of the compromise between memory usage and accuracy.
|
Analysis and Optimization of Loss Functions for Multiclass, Top-k, and
Multilabel Classification | Top-k error is currently a popular performance measure on large scale image
classification benchmarks such as ImageNet and Places. Despite its wide
acceptance, our understanding of this metric is limited as most of the previous
research is focused on its special case, the top-1 error. In this work, we
explore two directions that shed more light on the top-k error. First, we
provide an in-depth analysis of established and recently proposed single-label
multiclass methods along with a detailed account of efficient optimization
algorithms for them. Our results indicate that the softmax loss and the smooth
multiclass SVM are surprisingly competitive in top-k error uniformly across all
k, which can be explained by our analysis of multiclass top-k calibration.
Further improvements for a specific k are possible with a number of proposed
top-k loss functions. Second, we use the top-k methods to explore the
transition from multiclass to multilabel learning. In particular, we find that
it is possible to obtain effective multilabel classifiers on Pascal VOC using a
single label per image for training, while the gap between multiclass and
multilabel methods on MS COCO is more significant. Finally, our contribution of
efficient algorithms for training with the considered top-k and multilabel loss
functions is of independent interest.
|
Neurogenesis Deep Learning | Neural machine learning methods, such as deep neural networks (DNN), have
achieved remarkable success in a number of complex data processing tasks. These
methods have arguably had their strongest impact on tasks such as image and
audio processing - data processing domains in which humans have long held clear
advantages over conventional algorithms. In contrast to biological neural
systems, which are capable of learning continuously, deep artificial networks
have a limited ability for incorporating new information in an already trained
network. As a result, methods for continuous learning are potentially highly
impactful in enabling the application of deep networks to dynamic data sets.
Here, inspired by the process of adult neurogenesis in the hippocampus, we
explore the potential for adding new neurons to deep layers of artificial
neural networks in order to facilitate their acquisition of novel information
while preserving previously trained data representations. Our results on the
MNIST handwritten digit dataset and the NIST SD 19 dataset, which includes
lower and upper case letters and digits, demonstrate that neurogenesis is well
suited for addressing the stability-plasticity dilemma that has long challenged
adaptive machine learning algorithms.
|
Online Reinforcement Learning for Real-Time Exploration in Continuous
State and Action Markov Decision Processes | This paper presents a new method to learn online policies in continuous
state, continuous action, model-free Markov decision processes, with two
properties that are crucial for practical applications. First, the policies are
implementable with a very low computational cost: once the policy is computed,
the action corresponding to a given state is obtained in logarithmic time with
respect to the number of samples used. Second, our method is versatile: it does
not rely on any a priori knowledge of the structure of optimal policies. We
build upon the Fitted Q-iteration algorithm which represents the $Q$-value as
the average of several regression trees. Our algorithm, the Fitted Policy
Forest algorithm (FPF), computes a regression forest representing the Q-value
and transforms it into a single tree representing the policy, while keeping
control on the size of the policy using resampling and leaf merging. We
introduce an adaptation of Multi-Resolution Exploration (MRE) which is
particularly suited to FPF. We assess the performance of FPF on three classical
benchmarks for reinforcement learning: the "Inverted Pendulum", the "Double
Integrator" and "Car on the Hill" and show that FPF equals or outperforms other
algorithms, although these algorithms rely on the use of particular
representations of the policies, especially chosen in order to fit each of the
three problems. Finally, we exhibit that the combination of FPF and MRE allows
to find nearly optimal solutions in problems where $\epsilon$-greedy approaches
would fail.
|
A Unit Selection Methodology for Music Generation Using Deep Neural
Networks | Several methods exist for a computer to generate music based on data
including Markov chains, recurrent neural networks, recombinancy, and grammars.
We explore the use of unit selection and concatenation as a means of generating
music using a procedure based on ranking, where, we consider a unit to be a
variable length number of measures of music. We first examine whether a unit
selection method, that is restricted to a finite size unit library, can be
sufficient for encompassing a wide spectrum of music. We do this by developing
a deep autoencoder that encodes a musical input and reconstructs the input by
selecting from the library. We then describe a generative model that combines a
deep structured semantic model (DSSM) with an LSTM to predict the next unit,
where units consist of four, two, and one measures of music. We evaluate the
generative model using objective metrics including mean rank and accuracy and
with a subjective listening test in which expert musicians are asked to
complete a forced-choiced ranking task. We compare our model to a note-level
generative baseline that consists of a stacked LSTM trained to predict forward
by one note.
|
Generalizable Features From Unsupervised Learning | Humans learn a predictive model of the world and use this model to reason
about future events and the consequences of actions. In contrast to most
machine predictors, we exhibit an impressive ability to generalize to unseen
scenarios and reason intelligently in these settings. One important aspect of
this ability is physical intuition(Lake et al., 2016). In this work, we explore
the potential of unsupervised learning to find features that promote better
generalization to settings outside the supervised training distribution. Our
task is predicting the stability of towers of square blocks. We demonstrate
that an unsupervised model, trained to predict future frames of a video
sequence of stable and unstable block configurations, can yield features that
support extrapolating stability prediction to blocks configurations outside the
training set distribution
|
Tensor Decompositions via Two-Mode Higher-Order SVD (HOSVD) | Tensor decompositions have rich applications in statistics and machine
learning, and developing efficient, accurate algorithms for the problem has
received much attention recently. Here, we present a new method built on
Kruskal's uniqueness theorem to decompose symmetric, nearly orthogonally
decomposable tensors. Unlike the classical higher-order singular value
decomposition which unfolds a tensor along a single mode, we consider
unfoldings along two modes and use rank-1 constraints to characterize the
underlying components. This tensor decomposition method provably handles a
greater level of noise compared to previous methods and achieves a high
estimation accuracy. Numerical results demonstrate that our algorithm is robust
to various noise distributions and that it performs especially favorably as the
order increases.
|
Knowledge Completion for Generics using Guided Tensor Factorization | Given a knowledge base or KB containing (noisy) facts about common nouns or
generics, such as "all trees produce oxygen" or "some animals live in forests",
we consider the problem of inferring additional such facts at a precision
similar to that of the starting KB. Such KBs capture general knowledge about
the world, and are crucial for various applications such as question answering.
Different from commonly studied named entity KBs such as Freebase, generics KBs
involve quantification, have more complex underlying regularities, tend to be
more incomplete, and violate the commonly used locally closed world assumption
(LCWA). We show that existing KB completion methods struggle with this new
task, and present the first approach that is successful. Our results
demonstrate that external information, such as relation schemas and entity
taxonomies, if used appropriately, can be a surprisingly powerful tool in this
setting. First, our simple yet effective knowledge guided tensor factorization
approach achieves state-of-the-art results on two generics KBs (80% precise)
for science, doubling their size at 74%-86% precision. Second, our novel
taxonomy guided, submodular, active learning method for collecting annotations
about rare entities (e.g., oriole, a bird) is 6x more effective at inferring
further new facts about them than multiple active learning baselines.
|
Inverse Compositional Spatial Transformer Networks | In this paper, we establish a theoretical connection between the classical
Lucas & Kanade (LK) algorithm and the emerging topic of Spatial Transformer
Networks (STNs). STNs are of interest to the vision and learning communities
due to their natural ability to combine alignment and classification within the
same theoretical framework. Inspired by the Inverse Compositional (IC) variant
of the LK algorithm, we present Inverse Compositional Spatial Transformer
Networks (IC-STNs). We demonstrate that IC-STNs can achieve better performance
than conventional STNs with less model capacity; in particular, we show
superior performance in pure image alignment tasks as well as joint
alignment/classification problems on real-world problems.
|
Design of Data-Driven Mathematical Laws for Optimal Statistical
Classification Systems | This article will devise data-driven, mathematical laws that generate
optimal, statistical classification systems which achieve minimum error rates
for data distributions with unchanging statistics. Thereby, I will design
learning machines that minimize the expected risk or probability of
misclassification. I will devise a system of fundamental equations of binary
classification for a classification system in statistical equilibrium. I will
use this system of equations to formulate the problem of learning unknown,
linear and quadratic discriminant functions from data as a locus problem,
thereby formulating geometric locus methods within a statistical framework.
Solving locus problems involves finding equations of curves or surfaces defined
by given properties and finding graphs or loci of given equations. I will
devise three systems of data-driven, locus equations that generate optimal,
statistical classification systems. Each class of learning machines satisfies
fundamental statistical laws for a classification system in statistical
equilibrium. Thereby, I will formulate three classes of learning machines that
are scalable modules for optimal, statistical pattern recognition systems, all
of which are capable of performing a wide variety of statistical pattern
recognition tasks, where any given M-class statistical pattern recognition
system exhibits optimal generalization performance for an M-class feature
space.
|
Identification of release sources in advection-diffusion system by
machine learning combined with Green function inverse method | The identification of sources of advection-diffusion transport is based
usually on solving complex ill-posed inverse models against the available
state- variable data records. However, if there are several sources with
different locations and strengths, the data records represent mixtures rather
than the separate influences of the original sources. Importantly, the number
of these original release sources is typically unknown, which hinders
reliability of the classical inverse-model analyses. To address this challenge,
we present here a novel hybrid method for identification of the unknown number
of release sources. Our hybrid method, called HNMF, couples unsupervised
learning based on Nonnegative Matrix Factorization (NMF) and inverse-analysis
Green functions method. HNMF synergistically performs decomposition of the
recorded mixtures, finds the number of the unknown sources and uses the Green
function of advection-diffusion equation to identify their characteristics. In
the paper, we introduce the method and demonstrate that it is capable of
identifying the advection velocity and dispersivity of the medium as well as
the unknown number, locations, and properties of various sets of synthetic
release sources with different space and time dependencies, based only on the
recorded data. HNMF can be applied directly to any problem controlled by a
partial-differential parabolic equation where mixtures of an unknown number of
sources are measured at multiple locations.
|
Nonnegative Matrix Factorization for identification of unknown number of
sources emitting delayed signals | Factor analysis is broadly used as a powerful unsupervised machine learning
tool for reconstruction of hidden features in recorded mixtures of signals. In
the case of a linear approximation, the mixtures can be decomposed by a variety
of model-free Blind Source Separation (BSS) algorithms. Most of the available
BSS algorithms consider an instantaneous mixing of signals, while the case when
the mixtures are linear combinations of signals with delays is less explored.
Especially difficult is the case when the number of sources of the signals with
delays is unknown and has to be determined from the data as well. To address
this problem, in this paper, we present a new method based on Nonnegative
Matrix Factorization (NMF) that is capable of identifying: (a) the unknown
number of the sources, (b) the delays and speed of propagation of the signals,
and (c) the locations of the sources. Our method can be used to decompose
records of mixtures of signals with delays emitted by an unknown number of
sources in a nondispersive medium, based only on recorded data. This is the
case, for example, when electromagnetic signals from multiple antennas are
received asynchronously; or mixtures of acoustic or seismic signals recorded by
sensors located at different positions; or when a shift in frequency is induced
by the Doppler effect. By applying our method to synthetic datasets, we
demonstrate its ability to identify the unknown number of sources as well as
the waveforms, the delays, and the strengths of the signals. Using Bayesian
analysis, we also evaluate estimation uncertainties and identify the region of
likelihood where the positions of the sources can be found.
|
Hybrid Repeat/Multi-point Sampling for Highly Volatile Objective
Functions | A key drawback of the current generation of artificial decision-makers is
that they do not adapt well to changes in unexpected situations. This paper
addresses the situation in which an AI for aerial dog fighting, with tunable
parameters that govern its behavior, will optimize behavior with respect to an
objective function that must be evaluated and learned through simulations. Once
this objective function has been modeled, the agent can then choose its desired
behavior in different situations. Bayesian optimization with a Gaussian Process
surrogate is used as the method for investigating the objective function. One
key benefit is that during optimization the Gaussian Process learns a global
estimate of the true objective function, with predicted outcomes and a
statistical measure of confidence in areas that haven't been investigated yet.
However, standard Bayesian optimization does not perform consistently or
provide an accurate Gaussian Process surrogate function for highly volatile
objective functions. We treat these problems by introducing a novel sampling
technique called Hybrid Repeat/Multi-point Sampling. This technique gives the
AI ability to learn optimum behaviors in a highly uncertain environment. More
importantly, it not only improves the reliability of the optimization, but also
creates a better model of the entire objective surface. With this improved
model the agent is equipped to better adapt behaviors.
|
An empirical analysis of the optimization of deep network loss surfaces | The success of deep neural networks hinges on our ability to accurately and
efficiently optimize high-dimensional, non-convex functions. In this paper, we
empirically investigate the loss functions of state-of-the-art networks, and
how commonly-used stochastic gradient descent variants optimize these loss
functions. To do this, we visualize the loss function by projecting them down
to low-dimensional spaces chosen based on the convergence points of different
optimization algorithms. Our observations suggest that optimization algorithms
encounter and choose different descent directions at many saddle points to find
different final weights. Based on consistency we observe across re-runs of the
same stochastic optimization algorithm, we hypothesize that each optimization
algorithm makes characteristic choices at these saddle points.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.