title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Multilinear Low-Rank Tensors on Graphs & Applications | We propose a new framework for the analysis of low-rank tensors which lies at
the intersection of spectral graph theory and signal processing. As a first
step, we present a new graph based low-rank decomposition which approximates
the classical low-rank SVD for matrices and multi-linear SVD for tensors. Then,
building on this novel decomposition we construct a general class of convex
optimization problems for approximately solving low-rank tensor inverse
problems, such as tensor Robust PCA. The whole framework is named as
'Multilinear Low-rank tensors on Graphs (MLRTG)'. Our theoretical analysis
shows: 1) MLRTG stands on the notion of approximate stationarity of
multi-dimensional signals on graphs and 2) the approximation error depends on
the eigen gaps of the graphs. We demonstrate applications for a wide variety of
4 artificial and 12 real tensor datasets, such as EEG, FMRI, BCI, surveillance
videos and hyperspectral images. Generalization of the tensor concepts to
non-euclidean domain, orders of magnitude speed-up, low-memory requirement and
significantly enhanced performance at low SNR are the key aspects of our
framework.
|
The Power of Side-information in Subgraph Detection | In this work, we tackle the problem of hidden community detection. We
consider Belief Propagation (BP) applied to the problem of detecting a hidden
Erd\H{o}s-R\'enyi (ER) graph embedded in a larger and sparser ER graph, in the
presence of side-information. We derive two related algorithms based on BP to
perform subgraph detection in the presence of two kinds of side-information.
The first variant of side-information consists of a set of nodes, called cues,
known to be from the subgraph. The second variant of side-information consists
of a set of nodes that are cues with a given probability. It was shown in past
works that BP without side-information fails to detect the subgraph correctly
when an effective signal-to-noise ratio (SNR) parameter falls below a
threshold. In contrast, in the presence of non-trivial side-information, we
show that the BP algorithm achieves asymptotically zero error for any value of
the SNR parameter. We validate our results through simulations on synthetic
datasets as well as on a few real world networks.
|
Constrained Low-Rank Learning Using Least Squares-Based Regularization | Low-rank learning has attracted much attention recently due to its efficacy
in a rich variety of real-world tasks, e.g., subspace segmentation and image
categorization. Most low-rank methods are incapable of capturing
low-dimensional subspace for supervised learning tasks, e.g., classification
and regression. This paper aims to learn both the discriminant low-rank
representation (LRR) and the robust projecting subspace in a supervised manner.
To achieve this goal, we cast the problem into a constrained rank minimization
framework by adopting the least squares regularization. Naturally, the data
label structure tends to resemble that of the corresponding low-dimensional
representation, which is derived from the robust subspace projection of clean
data by low-rank learning. Moreover, the low-dimensional representation of
original data can be paired with some informative structure by imposing an
appropriate constraint, e.g., Laplacian regularizer. Therefore, we propose a
novel constrained LRR method. The objective function is formulated as a
constrained nuclear norm minimization problem, which can be solved by the
inexact augmented Lagrange multiplier algorithm. Extensive experiments on image
classification, human pose estimation, and robust face recovery have confirmed
the superiority of our method.
|
Audio Event and Scene Recognition: A Unified Approach using Strongly and
Weakly Labeled Data | In this paper we propose a novel learning framework called Supervised and
Weakly Supervised Learning where the goal is to learn simultaneously from
weakly and strongly labeled data. Strongly labeled data can be simply
understood as fully supervised data where all labeled instances are available.
In weakly supervised learning only data is weakly labeled which prevents one
from directly applying supervised learning methods. Our proposed framework is
motivated by the fact that a small amount of strongly labeled data can give
considerable improvement over only weakly supervised learning. The primary
problem domain focus of this paper is acoustic event and scene detection in
audio recordings. We first propose a naive formulation for leveraging labeled
data in both forms. We then propose a more general framework for Supervised and
Weakly Supervised Learning (SWSL). Based on this general framework, we propose
a graph based approach for SWSL. Our main method is based on manifold
regularization on graphs in which we show that the unified learning can be
formulated as a constraint optimization problem which can be solved by
iterative concave-convex procedure (CCCP). Our experiments show that our
proposed framework can address several concerns of audio content analysis using
weakly labeled data.
|
Unsupervised Learning with Truncated Gaussian Graphical Models | Gaussian graphical models (GGMs) are widely used for statistical modeling,
because of ease of inference and the ubiquitous use of the normal distribution
in practical approximations. However, they are also known for their limited
modeling abilities, due to the Gaussian assumption. In this paper, we introduce
a novel variant of GGMs, which relaxes the Gaussian restriction and yet admits
efficient inference. Specifically, we impose a bipartite structure on the GGM
and govern the hidden variables by truncated normal distributions. The
nonlinearity of the model is revealed by its connection to rectified linear
unit (ReLU) neural networks. Meanwhile, thanks to the bipartite structure and
appealing properties of truncated normals, we are able to train the models
efficiently using contrastive divergence. We consider three output constructs,
accounting for real-valued, binary and count data. We further extend the model
to deep constructions and show that deep models can be used for unsupervised
pre-training of rectifier neural networks. Extensive experimental results are
provided to validate the proposed models and demonstrate their superiority over
competing models.
|
Robust Semi-Supervised Graph Classifier Learning with Negative Edge
Weights | In a semi-supervised learning scenario, (possibly noisy) partially observed
labels are used as input to train a classifier, in order to assign labels to
unclassified samples. In this paper, we study this classifier learning problem
from a graph signal processing (GSP) perspective. Specifically, by viewing a
binary classifier as a piecewise constant graph-signal in a high-dimensional
feature space, we cast classifier learning as a signal restoration problem via
a classical maximum a posteriori (MAP) formulation. Unlike previous
graph-signal restoration works, we consider in addition edges with negative
weights that signify anti-correlation between samples. One unfortunate
consequence is that the graph Laplacian matrix $\mathbf{L}$ can be indefinite,
and previously proposed graph-signal smoothness prior $\mathbf{x}^T \mathbf{L}
\mathbf{x}$ for candidate signal $\mathbf{x}$ can lead to pathological
solutions. In response, we derive an optimal perturbation matrix
$\boldsymbol{\Delta}$ - based on a fast lower-bound computation of the minimum
eigenvalue of $\mathbf{L}$ via a novel application of the Haynsworth inertia
additivity formula---so that $\mathbf{L} + \boldsymbol{\Delta}$ is positive
semi-definite, resulting in a stable signal prior. Further, instead of forcing
a hard binary decision for each sample, we define the notion of generalized
smoothness on graph that promotes ambiguity in the classifier signal. Finally,
we propose an algorithm based on iterative reweighted least squares (IRLS) that
solves the posed MAP problem efficiently. Extensive simulation results show
that our proposed algorithm outperforms both SVM variants and graph-based
classifiers using positive-edge graphs noticeably.
|
Iterative Orthogonal Feature Projection for Diagnosing Bias in Black-Box
Models | Predictive models are increasingly deployed for the purpose of determining
access to services such as credit, insurance, and employment. Despite potential
gains in productivity and efficiency, several potential problems have yet to be
addressed, particularly the potential for unintentional discrimination. We
present an iterative procedure, based on orthogonal projection of input
attributes, for enabling interpretability of black-box predictive models.
Through our iterative procedure, one can quantify the relative dependence of a
black-box model on its input attributes.The relative significance of the inputs
to a predictive model can then be used to assess the fairness (or
discriminatory extent) of such a model.
|
Oracle Complexity of Second-Order Methods for Finite-Sum Problems | Finite-sum optimization problems are ubiquitous in machine learning, and are
commonly solved using first-order methods which rely on gradient computations.
Recently, there has been growing interest in \emph{second-order} methods, which
rely on both gradients and Hessians. In principle, second-order methods can
require much fewer iterations than first-order methods, and hold the promise
for more efficient algorithms. Although computing and manipulating Hessians is
prohibitive for high-dimensional problems in general, the Hessians of
individual functions in finite-sum problems can often be efficiently computed,
e.g. because they possess a low-rank structure. Can second-order information
indeed be used to solve such problems more efficiently? In this paper, we
provide evidence that the answer -- perhaps surprisingly -- is negative, at
least in terms of worst-case guarantees. However, we also discuss what
additional assumptions and algorithmic approaches might potentially circumvent
this negative result.
|
PixelVAE: A Latent Variable Model for Natural Images | Natural image modeling is a landmark challenge of unsupervised learning.
Variational Autoencoders (VAEs) learn a useful latent representation and model
global structure well but have difficulty capturing small details. PixelCNN
models details very well, but lacks a latent code and is difficult to scale for
capturing large structures. We present PixelVAE, a VAE model with an
autoregressive decoder based on PixelCNN. Our model requires very few expensive
autoregressive layers compared to PixelCNN and learns latent codes that are
more compressed than a standard VAE while still capturing most non-trivial
structure. Finally, we extend our model to a hierarchy of latent variables at
different scales. Our model achieves state-of-the-art performance on binarized
MNIST, competitive performance on 64x64 ImageNet, and high-quality samples on
the LSUN bedrooms dataset.
|
Probabilistic Failure Analysis in Model Validation & Verification | Automated fault localization is an important issue in model validation and
verification. It helps the end users in analyzing the origin of failure. In
this work, we show the early experiments with probabilistic analysis approaches
in fault localization. Inspired by the Kullback-Leibler Divergence from
Bayesian probabilistic theory, we propose a suspiciousness factor to compute
the fault contribution for the transitions in the reachability graph of model
checking, using which to rank the potential faulty transitions. To
automatically locate design faults in the simulation model of detailed design,
we propose to use the statistical model Hidden Markov Model (HMM), which
provides statistically identical information to component's real behavior. The
core of this method is a fault localization algorithm that gives out the set of
suspicious ranked faulty components and a backward algorithm that computes the
matching degree between the HMM and the simulation model to evaluate the
confidence degree of the localization conclusion.
|
Learning Dexterous Manipulation Policies from Experience and Imitation | We explore learning-based approaches for feedback control of a dexterous
five-finger hand performing non-prehensile manipulation. First, we learn local
controllers that are able to perform the task starting at a predefined initial
state. These controllers are constructed using trajectory optimization with
respect to locally-linear time-varying models learned directly from sensor
data. In some cases, we initialize the optimizer with human demonstrations
collected via teleoperation in a virtual environment. We demonstrate that such
controllers can perform the task robustly, both in simulation and on the
physical platform, for a limited range of initial conditions around the trained
starting state. We then consider two interpolation methods for generalizing to
a wider range of initial conditions: deep learning, and nearest neighbors. We
find that nearest neighbors achieve higher performance. Nevertheless, the
neural network has its advantages: it uses only tactile and proprioceptive
feedback but no visual feedback about the object (i.e. it performs the task
blind) and learns a time-invariant policy. In contrast, the nearest neighbors
method switches between time-varying local controllers based on the proximity
of initial object states sensed via motion capture. While both generalization
methods leave room for improvement, our work shows that (i) local
trajectory-based controllers for complex non-prehensile manipulation tasks can
be constructed from surprisingly small amounts of training data, and (ii)
collections of such controllers can be interpolated to form more global
controllers. Results are summarized in the supplementary video:
https://youtu.be/E0wmO6deqjo
|
Convergence rate of stochastic k-means | We analyze online \cite{BottouBengio} and mini-batch \cite{Sculley} $k$-means
variants. Both scale up the widely used $k$-means algorithm via stochastic
approximation, and have become popular for large-scale clustering and
unsupervised feature learning. We show, for the first time, that starting with
any initial solution, they converge to a "local optimum" at rate
$O(\frac{1}{t})$ (in terms of the $k$-means objective) under general
conditions. In addition, we show if the dataset is clusterable, when
initialized with a simple and scalable seeding algorithm, mini-batch $k$-means
converges to an optimal $k$-means solution at rate $O(\frac{1}{t})$ with high
probability. The $k$-means objective is non-convex and non-differentiable: we
exploit ideas from recent work on stochastic gradient descent for non-convex
problems \cite{ge:sgd_tensor, balsubramani13} by providing a novel
characterization of the trajectory of $k$-means algorithm on its solution
space, and circumvent the non-differentiability problem via geometric insights
about $k$-means update.
|
Machine Learning Approach for Skill Evaluation in Robotic-Assisted
Surgery | Evaluating surgeon skill has predominantly been a subjective task.
Development of objective methods for surgical skill assessment are of increased
interest. Recently, with technological advances such as robotic-assisted
minimally invasive surgery (RMIS), new opportunities for objective and
automated assessment frameworks have arisen. In this paper, we applied machine
learning methods to automatically evaluate performance of the surgeon in RMIS.
Six important movement features were used in the evaluation including
completion time, path length, depth perception, speed, smoothness and
curvature. Different classification methods applied to discriminate expert and
novice surgeons. We test our method on real surgical data for suturing task and
compare the classification result with the ground truth data (obtained by
manual labeling). The experimental results show that the proposed framework can
classify surgical skill level with relatively high accuracy of 85.7%. This
study demonstrates the ability of machine learning methods to automatically
classify expert and novice surgeons using movement features for different RMIS
tasks. Due to the simplicity and generalizability of the introduced
classification method, it is easy to implement in existing trainers.
|
S3Pool: Pooling with Stochastic Spatial Sampling | Feature pooling layers (e.g., max pooling) in convolutional neural networks
(CNNs) serve the dual purpose of providing increasingly abstract
representations as well as yielding computational savings in subsequent
convolutional layers. We view the pooling operation in CNNs as a two-step
procedure: first, a pooling window (e.g., $2\times 2$) slides over the feature
map with stride one which leaves the spatial resolution intact, and second,
downsampling is performed by selecting one pixel from each non-overlapping
pooling window in an often uniform and deterministic (e.g., top-left) manner.
Our starting point in this work is the observation that this regularly spaced
downsampling arising from non-overlapping windows, although intuitive from a
signal processing perspective (which has the goal of signal reconstruction), is
not necessarily optimal for \emph{learning} (where the goal is to generalize).
We study this aspect and propose a novel pooling strategy with stochastic
spatial sampling (S3Pool), where the regular downsampling is replaced by a more
general stochastic version. We observe that this general stochasticity acts as
a strong regularizer, and can also be seen as doing implicit data augmentation
by introducing distortions in the feature maps. We further introduce a
mechanism to control the amount of distortion to suit different datasets and
architectures. To demonstrate the effectiveness of the proposed approach, we
perform extensive experiments on several popular image classification
benchmarks, observing excellent improvements over baseline models. Experimental
code is available at https://github.com/Shuangfei/s3pool.
|
Training Spiking Deep Networks for Neuromorphic Hardware | We describe a method to train spiking deep networks that can be run using
leaky integrate-and-fire (LIF) neurons, achieving state-of-the-art results for
spiking LIF networks on five datasets, including the large ImageNet ILSVRC-2012
benchmark. Our method for transforming deep artificial neural networks into
spiking networks is scalable and works with a wide range of neural
nonlinearities. We achieve these results by softening the neural response
function, such that its derivative remains bounded, and by training the network
with noise to provide robustness against the variability introduced by spikes.
Our analysis shows that implementations of these networks on neuromorphic
hardware will be many times more power-efficient than the equivalent
non-spiking networks on traditional hardware.
|
A Semi-Markov Switching Linear Gaussian Model for Censored Physiological
Data | Critically ill patients in regular wards are vulnerable to unanticipated
clinical dete- rioration which requires timely transfer to the intensive care
unit (ICU). To allow for risk scoring and patient monitoring in such a setting,
we develop a novel Semi- Markov Switching Linear Gaussian Model (SSLGM) for the
inpatients' physiol- ogy. The model captures the patients' latent clinical
states and their corresponding observable lab tests and vital signs. We present
an efficient unsupervised learn- ing algorithm that capitalizes on the
informatively censored data in the electronic health records (EHR) to learn the
parameters of the SSLGM; the learned model is then used to assess the new
inpatients' risk for clinical deterioration in an online fashion, allowing for
timely ICU admission. Experiments conducted on a het- erogeneous cohort of
6,094 patients admitted to a large academic medical center show that the
proposed model significantly outperforms the currently deployed risk scores
such as Rothman index, MEWS, SOFA and APACHE.
|
Net-Trim: Convex Pruning of Deep Neural Networks with Performance
Guarantee | We introduce and analyze a new technique for model reduction for deep neural
networks. While large networks are theoretically capable of learning
arbitrarily complex models, overfitting and model redundancy negatively affects
the prediction accuracy and model variance. Our Net-Trim algorithm prunes
(sparsifies) a trained network layer-wise, removing connections at each layer
by solving a convex optimization program. This program seeks a sparse set of
weights at each layer that keeps the layer inputs and outputs consistent with
the originally trained model. The algorithms and associated analysis are
applicable to neural networks operating with the rectified linear unit (ReLU)
as the nonlinear activation. We present both parallel and cascade versions of
the algorithm. While the latter can achieve slightly simpler models with the
same generalization performance, the former can be computed in a distributed
manner. In both cases, Net-Trim significantly reduces the number of connections
in the network, while also providing enough regularization to slightly reduce
the generalization error. We also provide a mathematical analysis of the
consistency between the initial network and the retrained model. To analyze the
model sample complexity, we derive the general sufficient conditions for the
recovery of a sparse transform matrix. For a single layer taking independent
Gaussian random vectors of length $N$ as inputs, we show that if the network
response can be described using a maximum number of $s$ non-zero weights per
node, these weights can be learned from $\mathcal{O}(s\log N)$ samples.
|
Graph Learning from Data under Structural and Laplacian Constraints | Graphs are fundamental mathematical structures used in various fields to
represent data, signals and processes. In this paper, we propose a novel
framework for learning/estimating graphs from data. The proposed framework
includes (i) formulation of various graph learning problems, (ii) their
probabilistic interpretations and (iii) associated algorithms. Specifically,
graph learning problems are posed as estimation of graph Laplacian matrices
from some observed data under given structural constraints (e.g., graph
connectivity and sparsity level). From a probabilistic perspective, the
problems of interest correspond to maximum a posteriori (MAP) parameter
estimation of Gaussian-Markov random field (GMRF) models, whose precision
(inverse covariance) is a graph Laplacian matrix. For the proposed graph
learning problems, specialized algorithms are developed by incorporating the
graph Laplacian and structural constraints. The experimental results
demonstrate that the proposed algorithms outperform the current
state-of-the-art methods in terms of accuracy and computational efficiency.
|
Bayesian optimization of hyper-parameters in reservoir computing | We describe a method for searching the optimal hyper-parameters in reservoir
computing, which consists of a Gaussian process with Bayesian optimization. It
provides an alternative to other frequently used optimization methods such as
grid, random, or manual search. In addition to a set of optimal
hyper-parameters, the method also provides a probability distribution of the
cost function as a function of the hyper-parameters. We apply this method to
two types of reservoirs: nonlinear delay nodes and echo state networks. It
shows excellent performance on all considered benchmarks, either matching or
significantly surpassing results found in the literature. In general, the
algorithm achieves optimal results in fewer iterations when compared to other
optimization methods. We have optimized up to six hyper-parameters
simultaneously, which would have been infeasible using, e.g., grid search. Due
to its automated nature, this method significantly reduces the need for expert
knowledge when optimizing the hyper-parameters in reservoir computing. Existing
software libraries for Bayesian optimization, such as Spearmint, make the
implementation of the algorithm straightforward. A fork of the Spearmint
framework along with a tutorial on how to use it in practice is available at
https://bitbucket.org/uhasseltmachinelearning/spearmint/
|
Deep Variational Inference Without Pixel-Wise Reconstruction | Variational autoencoders (VAEs), that are built upon deep neural networks
have emerged as popular generative models in computer vision. Most of the work
towards improving variational autoencoders has focused mainly on making the
approximations to the posterior flexible and accurate, leading to tremendous
progress. However, there have been limited efforts to replace pixel-wise
reconstruction, which have known shortcomings. In this work, we use real-valued
non-volume preserving transformations (real NVP) to exactly compute the
conditional likelihood of the data given the latent distribution. We show that
a simple VAE with this form of reconstruction is competitive with complicated
VAE structures, on image modeling tasks. As part of our model, we develop
powerful conditional coupling layers that enable real NVP to learn with fewer
intermediate layers.
|
A Learning Scheme for Microgrid Islanding and Reconnection | This paper introduces a potential learning scheme that can dynamically
predict the stability of the reconnection of sub-networks to a main grid. As
the future electrical power systems tend towards smarter and greener
technology, the deployment of self sufficient networks, or microgrids, becomes
more likely. Microgrids may operate on their own or synchronized with the main
grid, thus control methods need to take into account islanding and reconnecting
of said networks. The ability to optimally and safely reconnect a portion of
the grid is not well understood and, as of now, limited to raw synchronization
between interconnection points. A support vector machine (SVM) leveraging
real-time data from phasor measurement units (PMUs) is proposed to predict in
real time whether the reconnection of a sub-network to the main grid would lead
to stability or instability. A dynamics simulator fed with pre-acquired system
parameters is used to create training data for the SVM in various operating
states. The classifier was tested on a variety of cases and operating points to
ensure diversity. Accuracies of approximately 85% were observed throughout most
conditions when making dynamic predictions of a given network.
|
Approximating Wisdom of Crowds using K-RBMs | An important way to make large training sets is to gather noisy labels from
crowds of non experts. We propose a method to aggregate noisy labels collected
from a crowd of workers or annotators. Eliciting labels is important in tasks
such as judging web search quality and rating products. Our method assumes that
labels are generated by a probability distribution over items and labels. We
formulate the method by drawing parallels between Gaussian Mixture Models
(GMMs) and Restricted Boltzmann Machines (RBMs) and show that the problem of
vote aggregation can be viewed as one of clustering. We use K-RBMs to perform
clustering. We finally show some empirical evaluations over real datasets.
|
Fast On-Line Kernel Density Estimation for Active Object Localization | A major goal of computer vision is to enable computers to interpret visual
situations---abstract concepts (e.g., "a person walking a dog," "a crowd
waiting for a bus," "a picnic") whose image instantiations are linked more by
their common spatial and semantic structure than by low-level visual
similarity. In this paper, we propose a novel method for prior learning and
active object localization for this kind of knowledge-driven search in static
images. In our system, prior situation knowledge is captured by a set of
flexible, kernel-based density estimations---a situation model---that represent
the expected spatial structure of the given situation. These estimations are
efficiently updated by information gained as the system searches for relevant
objects, allowing the system to use context as it is discovered to narrow the
search.
More specifically, at any given time in a run on a test image, our system
uses image features plus contextual information it has discovered to identify a
small subset of training images---an importance cluster---that is deemed most
similar to the given test image, given the context. This subset is used to
generate an updated situation model in an on-line fashion, using an efficient
multipole expansion technique.
As a proof of concept, we apply our algorithm to a highly varied and
challenging dataset consisting of instances of a "dog-walking" situation. Our
results support the hypothesis that dynamically-rendered, context-based
probability models can support efficient object localization in visual
situations. Moreover, our approach is general enough to be applied to diverse
machine learning paradigms requiring interpretable, probabilistic
representations generated from partially observed data.
|
DeepCas: an End-to-end Predictor of Information Cascades | Information cascades, effectively facilitated by most social network
platforms, are recognized as a major factor in almost every social success and
disaster in these networks. Can cascades be predicted? While many believe that
they are inherently unpredictable, recent work has shown that some key
properties of information cascades, such as size, growth, and shape, can be
predicted by a machine learning algorithm that combines many features. These
predictors all depend on a bag of hand-crafting features to represent the
cascade network and the global network structure. Such features, always
carefully and sometimes mysteriously designed, are not easy to extend or to
generalize to a different platform or domain.
Inspired by the recent successes of deep learning in multiple data mining
tasks, we investigate whether an end-to-end deep learning approach could
effectively predict the future size of cascades. Such a method automatically
learns the representation of individual cascade graphs in the context of the
global network structure, without hand-crafted features and heuristics. We find
that node embeddings fall short of predictive power, and it is critical to
learn the representation of a cascade graph as a whole. We present algorithms
that learn the representation of cascade graphs in an end-to-end manner, which
significantly improve the performance of cascade prediction over strong
baselines that include feature based methods, node embedding methods, and graph
kernel methods. Our results also provide interesting implications for cascade
prediction in general.
|
Fully-adaptive Feature Sharing in Multi-Task Networks with Applications
in Person Attribute Classification | Multi-task learning aims to improve generalization performance of multiple
prediction tasks by appropriately sharing relevant information across them. In
the context of deep neural networks, this idea is often realized by
hand-designed network architectures with layers that are shared across tasks
and branches that encode task-specific features. However, the space of possible
multi-task deep architectures is combinatorially large and often the final
architecture is arrived at by manual exploration of this space subject to
designer's bias, which can be both error-prone and tedious. In this work, we
propose a principled approach for designing compact multi-task deep learning
architectures. Our approach starts with a thin network and dynamically widens
it in a greedy manner during training using a novel criterion that promotes
grouping of similar tasks together. Our Extensive evaluation on person
attributes classification tasks involving facial and clothing attributes
suggests that the models produced by the proposed method are fast, compact and
can closely match or exceed the state-of-the-art accuracy from strong baselines
by much more expensive models.
|
Spectral Convolution Networks | Previous research has shown that computation of convolution in the frequency
domain provides a significant speedup versus traditional convolution network
implementations. However, this performance increase comes at the expense of
repeatedly computing the transform and its inverse in order to apply other
network operations such as activation, pooling, and dropout. We show,
mathematically, how convolution and activation can both be implemented in the
frequency domain using either the Fourier or Laplace transformation. The main
contributions are a description of spectral activation under the Fourier
transform and a further description of an efficient algorithm for computing
both convolution and activation under the Laplace transform. By computing both
the convolution and activation functions in the frequency domain, we can reduce
the number of transforms required, as well as reducing overall complexity. Our
description of a spectral activation function, together with existing spectral
analogs of other network functions may then be used to compose a fully spectral
implementation of a convolution network.
|
Reinforcement Learning with Unsupervised Auxiliary Tasks | Deep reinforcement learning agents have achieved state-of-the-art results by
directly maximising cumulative reward. However, environments contain a much
wider variety of possible training signals. In this paper, we introduce an
agent that also maximises many other pseudo-reward functions simultaneously by
reinforcement learning. All of these tasks share a common representation that,
like unsupervised learning, continues to develop in the absence of extrinsic
rewards. We also introduce a novel mechanism for focusing this representation
upon extrinsic rewards, so that learning can rapidly adapt to the most relevant
aspects of the actual task. Our agent significantly outperforms the previous
state-of-the-art on Atari, averaging 880\% expert human performance, and a
challenging suite of first-person, three-dimensional \emph{Labyrinth} tasks
leading to a mean speedup in learning of 10$\times$ and averaging 87\% expert
human performance on Labyrinth.
|
The ZipML Framework for Training Models with End-to-End Low Precision:
The Cans, the Cannots, and a Little Bit of Deep Learning | Recently there has been significant interest in training machine-learning
models at low precision: by reducing precision, one can reduce computation and
communication by one order of magnitude. We examine training at reduced
precision, both from a theoretical and practical perspective, and ask: is it
possible to train models at end-to-end low precision with provable guarantees?
Can this lead to consistent order-of-magnitude speedups? We present a framework
called ZipML to answer these questions. For linear models, the answer is yes.
We develop a simple framework based on one simple but novel strategy called
double sampling. Our framework is able to execute training at low precision
with no bias, guaranteeing convergence, whereas naive quantization would
introduce significant bias. We validate our framework across a range of
applications, and show that it enables an FPGA prototype that is up to 6.5x
faster than an implementation using full 32-bit precision. We further develop a
variance-optimal stochastic quantization strategy and show that it can make a
significant difference in a variety of settings. When applied to linear models
together with double sampling, we save up to another 1.7x in data movement
compared with uniform quantization. When training deep networks with quantized
models, we achieve higher accuracy than the state-of-the-art XNOR-Net. Finally,
we extend our framework through approximation to non-linear models, such as
SVM. We show that, although using low-precision data induces bias, we can
appropriately bound and control the bias. We find in practice 8-bit precision
is often sufficient to converge to the correct solution. Interestingly,
however, in practice we notice that our framework does not always outperform
the naive rounding approach. We discuss this negative result in detail.
|
Composing Music with Grammar Argumented Neural Networks and Note-Level
Encoding | Creating aesthetically pleasing pieces of art, including music, has been a
long-term goal for artificial intelligence research. Despite recent successes
of long-short term memory (LSTM) recurrent neural networks (RNNs) in sequential
learning, LSTM neural networks have not, by themselves, been able to generate
natural-sounding music conforming to music theory. To transcend this
inadequacy, we put forward a novel method for music composition that combines
the LSTM with Grammars motivated by music theory. The main tenets of music
theory are encoded as grammar argumented (GA) filters on the training data,
such that the machine can be trained to generate music inheriting the
naturalness of human-composed pieces from the original dataset while adhering
to the rules of music theory. Unlike previous approaches, pitches and durations
are encoded as one semantic entity, which we refer to as note-level encoding.
This allows easy implementation of music theory grammars, as well as closer
emulation of the thinking pattern of a musician. Although the GA rules are
applied to the training data and never directly to the LSTM music generation,
our machine still composes music that possess high incidences of diatonic scale
notes, small pitch intervals and chords, in deference to music theory.
|
Solving Cold-Start Problem in Large-scale Recommendation Engines: A Deep
Learning Approach | Collaborative Filtering (CF) is widely used in large-scale recommendation
engines because of its efficiency, accuracy and scalability. However, in
practice, the fact that recommendation engines based on CF require interactions
between users and items before making recommendations, make it inappropriate
for new items which haven't been exposed to the end users to interact with.
This is known as the cold-start problem. In this paper we introduce a novel
approach which employs deep learning to tackle this problem in any CF based
recommendation engine. One of the most important features of the proposed
technique is the fact that it can be applied on top of any existing CF based
recommendation engine without changing the CF core. We successfully applied
this technique to overcome the item cold-start problem in Careerbuilder's CF
based recommendation engine. Our experiments show that the proposed technique
is very efficient to resolve the cold-start problem while maintaining high
accuracy of the CF recommendations.
|
Algebraic multigrid support vector machines | The support vector machine is a flexible optimization-based technique widely
used for classification problems. In practice, its training part becomes
computationally expensive on large-scale data sets because of such reasons as
the complexity and number of iterations in parameter fitting methods,
underlying optimization solvers, and nonlinearity of kernels. We introduce a
fast multilevel framework for solving support vector machine models that is
inspired by the algebraic multigrid. Significant improvement in the running has
been achieved without any loss in the quality. The proposed technique is highly
beneficial on imbalanced sets. We demonstrate computational results on publicly
available and industrial data sets.
|
Robust Hashing for Multi-View Data: Jointly Learning Low-Rank Kernelized
Similarity Consensus and Hash Functions | Learning hash functions/codes for similarity search over multi-view data is
attracting increasing attention, where similar hash codes are assigned to the
data objects characterizing consistently neighborhood relationship across
views. Traditional methods in this category inherently suffer three
limitations: 1) they commonly adopt a two-stage scheme where similarity matrix
is first constructed, followed by a subsequent hash function learning; 2) these
methods are commonly developed on the assumption that data samples with
multiple representations are noise-free,which is not practical in real-life
applications; 3) they often incur cumbersome training model caused by the
neighborhood graph construction using all $N$ points in the database ($O(N)$).
In this paper, we motivate the problem of jointly and efficiently training the
robust hash functions over data objects with multi-feature representations
which may be noise corrupted. To achieve both the robustness and training
efficiency, we propose an approach to effectively and efficiently learning
low-rank kernelized \footnote{We use kernelized similarity rather than kernel,
as it is not a squared symmetric matrix for data-landmark affinity matrix.}
hash functions shared across views. Specifically, we utilize landmark graphs to
construct tractable similarity matrices in multi-views to automatically
discover neighborhood structure in the data. To learn robust hash functions, a
latent low-rank kernel function is used to construct hash functions in order to
accommodate linearly inseparable data. In particular, a latent kernelized
similarity matrix is recovered by rank minimization on multiple kernel-based
similarity matrices. Extensive experiments on real-world multi-view datasets
validate the efficacy of our method in the presence of error corruptions.
|
Automatic Node Selection for Deep Neural Networks using Group Lasso
Regularization | We examine the effect of the Group Lasso (gLasso) regularizer in selecting
the salient nodes of Deep Neural Network (DNN) hidden layers by applying a
DNN-HMM hybrid speech recognizer to TED Talks speech data. We test two types of
gLasso regularization, one for outgoing weight vectors and another for incoming
weight vectors, as well as two sizes of DNNs: 2048 hidden layer nodes and 4096
nodes. Furthermore, we compare gLasso and L2 regularizers. Our experiment
results demonstrate that our DNN training, in which the gLasso regularizer was
embedded, successfully selected the hidden layer nodes that are necessary and
sufficient for achieving high classification power.
|
DelugeNets: Deep Networks with Efficient and Flexible Cross-layer
Information Inflows | Deluge Networks (DelugeNets) are deep neural networks which efficiently
facilitate massive cross-layer information inflows from preceding layers to
succeeding layers. The connections between layers in DelugeNets are established
through cross-layer depthwise convolutional layers with learnable filters,
acting as a flexible yet efficient selection mechanism. DelugeNets can
propagate information across many layers with greater flexibility and utilize
network parameters more effectively compared to ResNets, whilst being more
efficient than DenseNets. Remarkably, a DelugeNet model with just model
complexity of 4.31 GigaFLOPs and 20.2M network parameters, achieve
classification errors of 3.76% and 19.02% on CIFAR-10 and CIFAR-100 dataset
respectively. Moreover, DelugeNet-122 performs competitively to ResNet-200 on
ImageNet dataset, despite costing merely half of the computations needed by the
latter.
|
Boosting Variational Inference | Variational inference (VI) provides fast approximations of a Bayesian
posterior in part because it formulates posterior approximation as an
optimization problem: to find the closest distribution to the exact posterior
over some family of distributions. For practical reasons, the family of
distributions in VI is usually constrained so that it does not include the
exact posterior, even as a limit point. Thus, no matter how long VI is run, the
resulting approximation will not approach the exact posterior. We propose to
instead consider a more flexible approximating family consisting of all
possible finite mixtures of a parametric base distribution (e.g., Gaussian).
For efficient inference, we borrow ideas from gradient boosting to develop an
algorithm we call boosting variational inference (BVI). BVI iteratively
improves the current approximation by mixing it with a new component from the
base distribution family and thereby yields progressively more accurate
posterior approximations as more computing time is spent. Unlike a number of
common VI variants including mean-field VI, BVI is able to capture
multimodality, general posterior covariance, and nonstandard posterior shapes.
|
Optical Flow Requires Multiple Strategies (but only one network) | We show that the matching problem that underlies optical flow requires
multiple strategies, depending on the amount of image motion and other factors.
We then study the implications of this observation on training a deep neural
network for representing image patches in the context of descriptor based
optical flow. We propose a metric learning method, which selects suitable
negative samples based on the nature of the true match. This type of training
produces a network that displays multiple strategies depending on the input and
leads to state of the art results on the KITTI 2012 and KITTI 2015 optical flow
benchmarks.
|
Inverting The Generator Of A Generative Adversarial Network | Generative adversarial networks (GANs) learn to synthesise new samples from a
high-dimensional distribution by passing samples drawn from a latent space
through a generative network. When the high-dimensional distribution describes
images of a particular data set, the network should learn to generate visually
similar image samples for latent variables that are close to each other in the
latent space. For tasks such as image retrieval and image classification, it
may be useful to exploit the arrangement of the latent space by projecting
images into it, and using this as a representation for discriminative tasks.
GANs often consist of multiple layers of non-linear computations, making them
very difficult to invert. This paper introduces techniques for projecting image
samples into the latent space using any pre-trained GAN, provided that the
computational graph is available. We evaluate these techniques on both MNIST
digits and Omniglot handwritten characters. In the case of MNIST digits, we
show that projections into the latent space maintain information about the
style and the identity of the digit. In the case of Omniglot characters, we
show that even characters from alphabets that have not been seen during
training may be projected well into the latent space; this suggests that this
approach may have applications in one-shot learning.
|
Study on Feature Subspace of Archetypal Emotions for Speech Emotion
Recognition | Feature subspace selection is an important part in speech emotion
recognition. Most of the studies are devoted to finding a feature subspace for
representing all emotions. However, some studies have indicated that the
features associated with different emotions are not exactly the same. Hence,
traditional methods may fail to distinguish some of the emotions with just one
global feature subspace. In this work, we propose a new divide and conquer idea
to solve the problem. First, the feature subspaces are constructed for all the
combinations of every two different emotions (emotion-pair). Bi-classifiers are
then trained on these feature subspaces respectively. The final emotion
recognition result is derived by the voting and competition method.
Experimental results demonstrate that the proposed method can get better
results than the traditional multi-classification method.
|
GENESIM: genetic extraction of a single, interpretable model | Models obtained by decision tree induction techniques excel in being
interpretable.However, they can be prone to overfitting, which results in a low
predictive performance. Ensemble techniques are able to achieve a higher
accuracy. However, this comes at a cost of losing interpretability of the
resulting model. This makes ensemble techniques impractical in applications
where decision support, instead of decision making, is crucial.
To bridge this gap, we present the GENESIM algorithm that transforms an
ensemble of decision trees to a single decision tree with an enhanced
predictive performance by using a genetic algorithm. We compared GENESIM to
prevalent decision tree induction and ensemble techniques using twelve publicly
available data sets. The results show that GENESIM achieves a better predictive
performance on most of these data sets than decision tree induction techniques
and a predictive performance in the same order of magnitude as the ensemble
techniques. Moreover, the resulting model of GENESIM has a very low complexity,
making it very interpretable, in contrast to ensemble techniques.
|
Unimodal Thompson Sampling for Graph-Structured Arms | We study, to the best of our knowledge, the first Bayesian algorithm for
unimodal Multi-Armed Bandit (MAB) problems with graph structure. In this
setting, each arm corresponds to a node of a graph and each edge provides a
relationship, unknown to the learner, between two nodes in terms of expected
reward. Furthermore, for any node of the graph there is a path leading to the
unique node providing the maximum expected reward, along which the expected
reward is monotonically increasing. Previous results on this setting describe
the behavior of frequentist MAB algorithms. In our paper, we design a Thompson
Sampling-based algorithm whose asymptotic pseudo-regret matches the lower bound
for the considered setting. We show that -as it happens in a wide number of
scenarios- Bayesian MAB algorithms dramatically outperform frequentist ones. In
particular, we provide a thorough experimental evaluation of the performance of
our and state-of-the-art algorithms as the properties of the graph vary.
|
Relational Multi-Manifold Co-Clustering | Co-clustering targets on grouping the samples (e.g., documents, users) and
the features (e.g., words, ratings) simultaneously. It employs the dual
relation and the bilateral information between the samples and features. In
many realworld applications, data usually reside on a submanifold of the
ambient Euclidean space, but it is nontrivial to estimate the intrinsic
manifold of the data space in a principled way. In this study, we focus on
improving the co-clustering performance via manifold ensemble learning, which
is able to maximally approximate the intrinsic manifolds of both the sample and
feature spaces. To achieve this, we develop a novel co-clustering algorithm
called Relational Multi-manifold Co-clustering (RMC) based on symmetric
nonnegative matrix tri-factorization, which decomposes the relational data
matrix into three submatrices. This method considers the intertype relationship
revealed by the relational data matrix, and also the intra-type information
reflected by the affinity matrices encoded on the sample and feature data
distributions. Specifically, we assume the intrinsic manifold of the sample or
feature space lies in a convex hull of some pre-defined candidate manifolds. We
want to learn a convex combination of them to maximally approach the desired
intrinsic manifold. To optimize the objective function, the multiplicative
rules are utilized to update the submatrices alternatively. Besides, both the
entropic mirror descent algorithm and the coordinate descent algorithm are
exploited to learn the manifold coefficient vector. Extensive experiments on
documents, images and gene expression data sets have demonstrated the
superiority of the proposed algorithm compared to other well-established
methods.
|
A Multi-Modal Graph-Based Semi-Supervised Pipeline for Predicting Cancer
Survival | Cancer survival prediction is an active area of research that can help
prevent unnecessary therapies and improve patient's quality of life. Gene
expression profiling is being widely used in cancer studies to discover
informative biomarkers that aid predict different clinical endpoint prediction.
We use multiple modalities of data derived from RNA deep-sequencing (RNA-seq)
to predict survival of cancer patients. Despite the wealth of information
available in expression profiles of cancer tumors, fulfilling the
aforementioned objective remains a big challenge, for the most part, due to the
paucity of data samples compared to the high dimension of the expression
profiles. As such, analysis of transcriptomic data modalities calls for
state-of-the-art big-data analytics techniques that can maximally use all the
available data to discover the relevant information hidden within a significant
amount of noise. In this paper, we propose a pipeline that predicts cancer
patients' survival by exploiting the structure of the input (manifold learning)
and by leveraging the unlabeled samples using Laplacian support vector
machines, a graph-based semi supervised learning (GSSL) paradigm. We show that
under certain circumstances, no single modality per se will result in the best
accuracy and by fusing different models together via a stacked generalization
strategy, we may boost the accuracy synergistically. We apply our approach to
two cancer datasets and present promising results. We maintain that a similar
pipeline can be used for predictive tasks where labeled samples are expensive
to acquire.
|
Learning to reinforcement learn | In recent years deep reinforcement learning (RL) systems have attained
superhuman performance in a number of challenging task domains. However, a
major limitation of such applications is their demand for massive amounts of
training data. A critical present objective is thus to develop deep RL methods
that can adapt rapidly to new tasks. In the present work we introduce a novel
approach to this challenge, which we refer to as deep meta-reinforcement
learning. Previous work has shown that recurrent networks can support
meta-learning in a fully supervised context. We extend this approach to the RL
setting. What emerges is a system that is trained using one RL algorithm, but
whose recurrent dynamics implement a second, quite separate RL procedure. This
second, learned RL algorithm can differ from the original one in arbitrary
ways. Importantly, because it is learned, it is configured to exploit structure
in the training domain. We unpack these points in a series of seven
proof-of-concept experiments, each of which examines a key aspect of deep
meta-RL. We consider prospects for extending and scaling up the approach, and
also point out some potentially important implications for neuroscience.
|
Gap Safe screening rules for sparsity enforcing penalties | In high dimensional regression settings, sparsity enforcing penalties have
proved useful to regularize the data-fitting term. A recently introduced
technique called screening rules propose to ignore some variables in the
optimization leveraging the expected sparsity of the solutions and consequently
leading to faster solvers. When the procedure is guaranteed not to discard
variables wrongly the rules are said to be safe. In this work, we propose a
unifying framework for generalized linear models regularized with standard
sparsity enforcing penalties such as $\ell_1$ or $\ell_1/\ell_2$ norms. Our
technique allows to discard safely more variables than previously considered
safe rules, particularly for low regularization parameters. Our proposed Gap
Safe rules (so called because they rely on duality gap computation) can cope
with any iterative solver but are particularly well suited to (block)
coordinate descent methods. Applied to many standard learning tasks, Lasso,
Sparse-Group Lasso, multi-task Lasso, binary and multinomial logistic
regression, etc., we report significant speed-ups compared to previously
proposed safe rules on all tested data sets.
|
Data Science in Service of Performing Arts: Applying Machine Learning to
Predicting Audience Preferences | Performing arts organizations aim to enrich their communities through the
arts. To do this, they strive to match their performance offerings to the taste
of those communities. Success relies on understanding audience preference and
predicting their behavior. Similar to most e-commerce or digital entertainment
firms, arts presenters need to recommend the right performance to the right
customer at the right time. As part of the Michigan Data Science Team (MDST),
we partnered with the University Musical Society (UMS), a non-profit performing
arts presenter housed in the University of Michigan, Ann Arbor. We are
providing UMS with analysis and business intelligence, utilizing historical
individual-level sales data. We built a recommendation system based on
collaborative filtering, gaining insights into the artistic preferences of
customers, along with the similarities between performances. To better
understand audience behavior, we used statistical methods from customer-base
analysis. We characterized customer heterogeneity via segmentation, and we
modeled customer cohorts to understand and predict ticket purchasing patterns.
Finally, we combined statistical modeling with natural language processing
(NLP) to explore the impact of wording in program descriptions. These ongoing
efforts provide a platform to launch targeted marketing campaigns, helping UMS
carry out its mission by allocating its resources more efficiently. Celebrating
its 138th season, UMS is a 2014 recipient of the National Medal of Arts, and it
continues to enrich communities by connecting world-renowned artists with
diverse audiences, especially students in their formative years. We aim to
contribute to that mission through data science and customer analytics.
|
Nothing Else Matters: Model-Agnostic Explanations By Identifying
Prediction Invariance | At the core of interpretable machine learning is the question of whether
humans are able to make accurate predictions about a model's behavior. Assumed
in this question are three properties of the interpretable output: coverage,
precision, and effort. Coverage refers to how often humans think they can
predict the model's behavior, precision to how accurate humans are in those
predictions, and effort is either the up-front effort required in interpreting
the model, or the effort required to make predictions about a model's behavior.
In this work, we propose anchor-LIME (aLIME), a model-agnostic technique that
produces high-precision rule-based explanations for which the coverage
boundaries are very clear. We compare aLIME to linear LIME with simulated
experiments, and demonstrate the flexibility of aLIME with qualitative examples
from a variety of domains and tasks.
|
Towards a Mathematical Understanding of the Difficulty in Learning with
Feedforward Neural Networks | Training deep neural networks for solving machine learning problems is one
great challenge in the field, mainly due to its associated optimisation problem
being highly non-convex. Recent developments have suggested that many training
algorithms do not suffer from undesired local minima under certain scenario,
and consequently led to great efforts in pursuing mathematical explanations for
such observations. This work provides an alternative mathematical understanding
of the challenge from a smooth optimisation perspective. By assuming exact
learning of finite samples, sufficient conditions are identified via a critical
point analysis to ensure any local minimum to be globally minimal as well.
Furthermore, a state of the art algorithm, known as the Generalised
Gauss-Newton (GGN) algorithm, is rigorously revisited as an approximate
Newton's algorithm, which shares the property of being locally quadratically
convergent to a global minimum under the condition of exact learning.
|
Associative Memories to Accelerate Approximate Nearest Neighbor Search | Nearest neighbor search is a very active field in machine learning for it
appears in many application cases, including classification and object
retrieval. In its canonical version, the complexity of the search is linear
with both the dimension and the cardinal of the collection of vectors the
search is performed in. Recently many works have focused on reducing the
dimension of vectors using quantization techniques or hashing, while providing
an approximate result. In this paper we focus instead on tackling the cardinal
of the collection of vectors. Namely, we introduce a technique that partitions
the collection of vectors and stores each part in its own associative memory.
When a query vector is given to the system, associative memories are polled to
identify which one contain the closest match. Then an exhaustive search is
conducted only on the part of vectors stored in the selected associative
memory. We study the effectiveness of the system when messages to store are
generated from i.i.d. uniform $\pm$1 random variables or 0-1 sparse i.i.d.
random variables. We also conduct experiment on both synthetic data and real
data and show it is possible to achieve interesting trade-offs between
complexity and accuracy.
|
"Influence Sketching": Finding Influential Samples In Large-Scale
Regressions | There is an especially strong need in modern large-scale data analysis to
prioritize samples for manual inspection. For example, the inspection could
target important mislabeled samples or key vulnerabilities exploitable by an
adversarial attack. In order to solve the "needle in the haystack" problem of
which samples to inspect, we develop a new scalable version of Cook's distance,
a classical statistical technique for identifying samples which unusually
strongly impact the fit of a regression model (and its downstream predictions).
In order to scale this technique up to very large and high-dimensional
datasets, we introduce a new algorithm which we call "influence sketching."
Influence sketching embeds random projections within the influence computation;
in particular, the influence score is calculated using the randomly projected
pseudo-dataset from the post-convergence Generalized Linear Model (GLM). We
validate that influence sketching can reliably and successfully discover
influential samples by applying the technique to a malware detection dataset of
over 2 million executable files, each represented with almost 100,000 features.
For example, we find that randomly deleting approximately 10% of training
samples reduces predictive accuracy only slightly from 99.47% to 99.45%,
whereas deleting the same number of samples with high influence sketch scores
reduces predictive accuracy all the way down to 90.24%. Moreover, we find that
influential samples are especially likely to be mislabeled. In the case study,
we manually inspect the most influential samples, and find that influence
sketching pointed us to new, previously unidentified pieces of malware.
|
Increasing the Interpretability of Recurrent Neural Networks Using
Hidden Markov Models | As deep neural networks continue to revolutionize various application
domains, there is increasing interest in making these powerful models more
understandable and interpretable, and narrowing down the causes of good and bad
predictions. We focus on recurrent neural networks, state of the art models in
speech recognition and translation. Our approach to increasing interpretability
is by combining a long short-term memory (LSTM) model with a hidden Markov
model (HMM), a simpler and more transparent model. We add the HMM state
probabilities to the output layer of the LSTM, and then train the HMM and LSTM
either sequentially or jointly. The LSTM can make use of the information from
the HMM, and fill in the gaps when the HMM is not performing well. A small
hybrid model usually performs better than a standalone LSTM of the same size,
especially on smaller data sets. We test the algorithms on text data and
medical time series data, and find that the LSTM and HMM learn complementary
information about the features in the text.
|
Analysis of a Design Pattern for Teaching with Features and Labels | We study the task of teaching a machine to classify objects using features
and labels. We introduce the Error-Driven-Featuring design pattern for teaching
using features and labels in which a teacher prefers to introduce features only
if they are needed. We analyze the potential risks and benefits of this
teaching pattern through the use of teaching protocols, illustrative examples,
and by providing bounds on the effort required for an optimal machine teacher
using a linear learning algorithm, the most commonly used type of learners in
interactive machine learning systems. Our analysis provides a deeper
understanding of potential trade-offs of using different learning algorithms
and between the effort required for featuring (creating new features) and
labeling (providing labels for objects).
|
A Characterization of Prediction Errors | Understanding prediction errors and determining how to fix them is critical
to building effective predictive systems. In this paper, we delineate four
types of prediction errors and demonstrate that these four types characterize
all prediction errors. In addition, we describe potential remedies and tools
that can be used to reduce the uncertainty when trying to determine the source
of a prediction error and when trying to take action to remove a prediction
errors.
|
Robust and Scalable Column/Row Sampling from Corrupted Big Data | Conventional sampling techniques fall short of drawing descriptive sketches
of the data when the data is grossly corrupted as such corruptions break the
low rank structure required for them to perform satisfactorily. In this paper,
we present new sampling algorithms which can locate the informative columns in
presence of severe data corruptions. In addition, we develop new scalable
randomized designs of the proposed algorithms. The proposed approach is
simultaneously robust to sparse corruption and outliers and substantially
outperforms the state-of-the-art robust sampling algorithms as demonstrated by
experiments conducted using both real and synthetic data.
|
Monte Carlo Tableau Proof Search | We study Monte Carlo Tree Search to guide proof search in tableau calculi.
This includes proposing a number of proof-state evaluation heuristics, some of
which are learnt from previous proofs. We present an implementation based on
the leanCoP prover. The system is trained and evaluated on a large suite of
related problems coming from the Mizar proof assistant, showing that it is
capable to find new and different proofs.
|
A Generalized Stochastic Variational Bayesian Hyperparameter Learning
Framework for Sparse Spectrum Gaussian Process Regression | While much research effort has been dedicated to scaling up sparse Gaussian
process (GP) models based on inducing variables for big data, little attention
is afforded to the other less explored class of low-rank GP approximations that
exploit the sparse spectral representation of a GP kernel. This paper presents
such an effort to advance the state of the art of sparse spectrum GP models to
achieve competitive predictive performance for massive datasets. Our
generalized framework of stochastic variational Bayesian sparse spectrum GP
(sVBSSGP) models addresses their shortcomings by adopting a Bayesian treatment
of the spectral frequencies to avoid overfitting, modeling these frequencies
jointly in its variational distribution to enable their interaction a
posteriori, and exploiting local data for boosting the predictive performance.
However, such structural improvements result in a variational lower bound that
is intractable to be optimized. To resolve this, we exploit a variational
parameterization trick to make it amenable to stochastic optimization.
Interestingly, the resulting stochastic gradient has a linearly decomposable
structure that can be exploited to refine our stochastic optimization method to
incur constant time per iteration while preserving its property of being an
unbiased estimator of the exact gradient of the variational lower bound.
Empirical evaluation on real-world datasets shows that sVBSSGP outperforms
state-of-the-art stochastic implementations of sparse GP models.
|
Faster variational inducing input Gaussian process classification | Gaussian processes (GP) provide a prior over functions and allow finding
complex regularities in data. Gaussian processes are successfully used for
classification/regression problems and dimensionality reduction. In this work
we consider the classification problem only. The complexity of standard methods
for GP-classification scales cubically with the size of the training dataset.
This complexity makes them inapplicable to big data problems. Therefore, a
variety of methods were introduced to overcome this limitation. In the paper we
focus on methods based on so called inducing inputs. This approach is based on
variational inference and proposes a particular lower bound for marginal
likelihood (evidence). This bound is then maximized w.r.t. parameters of kernel
function of the Gaussian process, thus fitting the model to data. The
computational complexity of this method is $O(nm^2)$, where $m$ is the number
of inducing inputs used by the model and is assumed to be substantially smaller
than the size of the dataset $n$. Recently, a new evidence lower bound for
GP-classification problem was introduced. It allows using stochastic
optimization, which makes it suitable for big data problems. However, the new
lower bound depends on $O(m^2)$ variational parameter, which makes optimization
challenging in case of big m. In this work we develop a new approach for
training inducing input GP models for classification problems. Here we use
quadratic approximation of several terms in the aforementioned evidence lower
bound, obtaining analytical expressions for optimal values of most of the
parameters in the optimization, thus sufficiently reducing the dimension of
optimization space. In our experiments we achieve as well or better results,
compared to the existing method. Moreover, our method doesn't require the user
to manually set the learning rate, making it more practical, than the existing
method.
|
Compacting Neural Network Classifiers via Dropout Training | We introduce dropout compaction, a novel method for training feed-forward
neural networks which realizes the performance gains of training a large model
with dropout regularization, yet extracts a compact neural network for run-time
efficiency. In the proposed method, we introduce a sparsity-inducing prior on
the per unit dropout retention probability so that the optimizer can
effectively prune hidden units during training. By changing the prior
hyperparameters, we can control the size of the resulting network. We performed
a systematic comparison of dropout compaction and competing methods on several
real-world speech recognition tasks and found that dropout compaction achieved
comparable accuracy with fewer than 50% of the hidden units, translating to a
2.5x speedup in run-time.
|
Learning Interpretability for Visualizations using Adapted Cox Models
through a User Experiment | In order to be useful, visualizations need to be interpretable. This paper
uses a user-based approach to combine and assess quality measures in order to
better model user preferences. Results show that cluster separability measures
are outperformed by a neighborhood conservation measure, even though the former
are usually considered as intuitively representative of user motives. Moreover,
combining measures, as opposed to using a single measure, further improves
prediction performances.
|
Variable Computation in Recurrent Neural Networks | Recurrent neural networks (RNNs) have been used extensively and with
increasing success to model various types of sequential data. Much of this
progress has been achieved through devising recurrent units and architectures
with the flexibility to capture complex statistics in the data, such as long
range dependency or localized attention phenomena. However, while many
sequential data (such as video, speech or language) can have highly variable
information flow, most recurrent models still consume input features at a
constant rate and perform a constant number of computations per time step,
which can be detrimental to both speed and model capacity. In this paper, we
explore a modification to existing recurrent units which allows them to learn
to vary the amount of computation they perform at each step, without prior
knowledge of the sequence's time structure. We show experimentally that not
only do our models require fewer operations, they also lead to better
performance overall on evaluation tasks.
|
Visualizing and Understanding Curriculum Learning for Long Short-Term
Memory Networks | Curriculum Learning emphasizes the order of training instances in a
computational learning setup. The core hypothesis is that simpler instances
should be learned early as building blocks to learn more complex ones. Despite
its usefulness, it is still unknown how exactly the internal representation of
models are affected by curriculum learning. In this paper, we study the effect
of curriculum learning on Long Short-Term Memory (LSTM) networks, which have
shown strong competency in many Natural Language Processing (NLP) problems. Our
experiments on sentiment analysis task and a synthetic task similar to sequence
prediction tasks in NLP show that curriculum learning has a positive effect on
the LSTM's internal states by biasing the model towards building constructive
representations i.e. the internal representation at the previous timesteps are
used as building blocks for the final prediction. We also find that smaller
models significantly improves when they are trained with curriculum learning.
Lastly, we show that curriculum learning helps more when the amount of training
data is limited.
|
GaDei: On Scale-up Training As A Service For Deep Learning | Deep learning (DL) training-as-a-service (TaaS) is an important emerging
industrial workload. The unique challenge of TaaS is that it must satisfy a
wide range of customers who have no experience and resources to tune DL
hyper-parameters, and meticulous tuning for each user's dataset is
prohibitively expensive. Therefore, TaaS hyper-parameters must be fixed with
values that are applicable to all users. IBM Watson Natural Language Classifier
(NLC) service, the most popular IBM cognitive service used by thousands of
enterprise-level clients around the globe, is a typical TaaS service. By
evaluating the NLC workloads, we show that only the conservative
hyper-parameter setup (e.g., small mini-batch size and small learning rate) can
guarantee acceptable model accuracy for a wide range of customers. We further
justify theoretically why such a setup guarantees better model convergence in
general. Unfortunately, the small mini-batch size causes a high volume of
communication traffic in a parameter-server based system. We characterize the
high communication bandwidth requirement of TaaS using representative
industrial deep learning workloads and demonstrate that none of the
state-of-the-art scale-up or scale-out solutions can satisfy such a
requirement. We then present GaDei, an optimized shared-memory based scale-up
parameter server design. We prove that the designed protocol is deadlock-free
and it processes each gradient exactly once. Our implementation is evaluated on
both commercial benchmarks and public benchmarks to demonstrate that it
significantly outperforms the state-of-the-art parameter-server based
implementation while maintaining the required accuracy and our implementation
reaches near the best possible runtime performance, constrained only by the
hardware limitation. Furthermore, to the best of our knowledge, GaDei is the
only scale-up DL system that provides fault-tolerance.
|
Foundations of Structural Causal Models with Cycles and Latent Variables | Structural causal models (SCMs), also known as (nonparametric) structural
equation models (SEMs), are widely used for causal modeling purposes. In
particular, acyclic SCMs, also known as recursive SEMs, form a well-studied
subclass of SCMs that generalize causal Bayesian networks to allow for latent
confounders. In this paper, we investigate SCMs in a more general setting,
allowing for the presence of both latent confounders and cycles. We show that
in the presence of cycles, many of the convenient properties of acyclic SCMs do
not hold in general: they do not always have a solution; they do not always
induce unique observational, interventional and counterfactual distributions; a
marginalization does not always exist, and if it exists the marginal model does
not always respect the latent projection; they do not always satisfy a Markov
property; and their graphs are not always consistent with their causal
semantics. We prove that for SCMs in general each of these properties does hold
under certain solvability conditions. Our work generalizes results for SCMs
with cycles that were only known for certain special cases so far. We introduce
the class of simple SCMs that extends the class of acyclic SCMs to the cyclic
setting, while preserving many of the convenient properties of acyclic SCMs.
With this paper we aim to provide the foundations for a general theory of
statistical causal modeling with SCMs.
|
Approximate Near Neighbors for General Symmetric Norms | We show that every symmetric normed space admits an efficient nearest
neighbor search data structure with doubly-logarithmic approximation.
Specifically, for every $n$, $d = n^{o(1)}$, and every $d$-dimensional
symmetric norm $\|\cdot\|$, there exists a data structure for
$\mathrm{poly}(\log \log n)$-approximate nearest neighbor search over
$\|\cdot\|$ for $n$-point datasets achieving $n^{o(1)}$ query time and
$n^{1+o(1)}$ space. The main technical ingredient of the algorithm is a
low-distortion embedding of a symmetric norm into a low-dimensional iterated
product of top-$k$ norms.
We also show that our techniques cannot be extended to general norms.
|
Using LSTM recurrent neural networks for monitoring the LHC
superconducting magnets | The superconducting LHC magnets are coupled with an electronic monitoring
system which records and analyses voltage time series reflecting their
performance. A currently used system is based on a range of preprogrammed
triggers which launches protection procedures when a misbehavior of the magnets
is detected. All the procedures used in the protection equipment were designed
and implemented according to known working scenarios of the system and are
updated and monitored by human operators.
This paper proposes a novel approach to monitoring and fault protection of
the Large Hadron Collider (LHC) superconducting magnets which employs
state-of-the-art Deep Learning algorithms. Consequently, the authors of the
paper decided to examine the performance of LSTM recurrent neural networks for
modeling of voltage time series of the magnets. In order to address this
challenging task different network architectures and hyper-parameters were used
to achieve the best possible performance of the solution. The regression
results were measured in terms of RMSE for different number of future steps and
history length taken into account for the prediction. The best result of
RMSE=0.00104 was obtained for a network of 128 LSTM cells within the internal
layer and 16 steps history buffer.
|
Spikes as regularizers | We present a confidence-based single-layer feed-forward learning algorithm
SPIRAL (Spike Regularized Adaptive Learning) relying on an encoding of
activation spikes. We adaptively update a weight vector relying on confidence
estimates and activation offsets relative to previous activity. We regularize
updates proportionally to item-level confidence and weight-specific support,
loosely inspired by the observation from neurophysiology that high spike rates
are sometimes accompanied by low temporal precision. Our experiments suggest
that the new learning algorithm SPIRAL is more robust and less prone to
overfitting than both the averaged perceptron and AROW.
|
Reinforcement Learning through Asynchronous Advantage Actor-Critic on a
GPU | We introduce a hybrid CPU/GPU version of the Asynchronous Advantage
Actor-Critic (A3C) algorithm, currently the state-of-the-art method in
reinforcement learning for various gaming tasks. We analyze its computational
traits and concentrate on aspects critical to leveraging the GPU's
computational power. We introduce a system of queues and a dynamic scheduling
strategy, potentially helpful for other asynchronous algorithms as well. Our
hybrid CPU/GPU version of A3C, based on TensorFlow, achieves a significant
speed up compared to a CPU implementation; we make it publicly available to
other researchers at https://github.com/NVlabs/GA3C .
|
Deep Clustering and Conventional Networks for Music Separation: Stronger
Together | Deep clustering is the first method to handle general audio separation
scenarios with multiple sources of the same type and an arbitrary number of
sources, performing impressively in speaker-independent speech separation
tasks. However, little is known about its effectiveness in other challenging
situations such as music source separation. Contrary to conventional networks
that directly estimate the source signals, deep clustering generates an
embedding for each time-frequency bin, and separates sources by clustering the
bins in the embedding space. We show that deep clustering outperforms
conventional networks on a singing voice separation task, in both matched and
mismatched conditions, even though conventional networks have the advantage of
end-to-end training for best signal approximation, presumably because its more
flexible objective engenders better regularization. Since the strengths of deep
clustering and conventional network architectures appear complementary, we
explore combining them in a single hybrid network trained via an approach akin
to multi-task learning. Remarkably, the combination significantly outperforms
either of its components.
|
Cross-model convolutional neural network for multiple modality data
representation | A novel data representation method of convolutional neural net- work (CNN) is
proposed in this paper to represent data of different modalities. We learn a
CNN model for the data of each modality to map the data of differ- ent
modalities to a common space, and regularize the new representations in the
common space by a cross-model relevance matrix. We further impose that the
class label of data points can also be predicted from the CNN representa- tions
in the common space. The learning problem is modeled as a minimiza- tion
problem, which is solved by an augmented Lagrange method (ALM) with updating
rules of Alternating direction method of multipliers (ADMM). The experiments
over benchmark of sequence data of multiple modalities show its advantage.
|
Local minima in training of neural networks | There has been a lot of recent interest in trying to characterize the error
surface of deep models. This stems from a long standing question. Given that
deep networks are highly nonlinear systems optimized by local gradient methods,
why do they not seem to be affected by bad local minima? It is widely believed
that training of deep models using gradient methods works so well because the
error surface either has no local minima, or if they exist they need to be
close in value to the global minimum. It is known that such results hold under
very strong assumptions which are not satisfied by real models. In this paper
we present examples showing that for such theorem to be true additional
assumptions on the data, initialization schemes and/or the model classes have
to be made. We look at the particular case of finite size datasets. We
demonstrate that in this scenario one can construct counter-examples (datasets
or initialization schemes) when the network does become susceptible to bad
local minima over the weight space.
|
Learning the Number of Neurons in Deep Networks | Nowadays, the number of layers and of neurons in each layer of a deep network
are typically set manually. While very deep and wide networks have proven
effective in general, they come at a high memory and computation cost, thus
making them impractical for constrained platforms. These networks, however, are
known to have many redundant parameters, and could thus, in principle, be
replaced by more compact architectures. In this paper, we introduce an approach
to automatically determining the number of neurons in each layer of a deep
network during learning. To this end, we propose to make use of structured
sparsity during learning. More precisely, we use a group sparsity regularizer
on the parameters of the network, where each group is defined to act on a
single neuron. Starting from an overcomplete network, we show that our approach
can reduce the number of parameters by up to 80\% while retaining or even
improving the network accuracy.
|
Quantized neural network design under weight capacity constraint | The complexity of deep neural network algorithms for hardware implementation
can be lowered either by scaling the number of units or reducing the
word-length of weights. Both approaches, however, can accompany the performance
degradation although many types of research are conducted to relieve this
problem. Thus, it is an important question which one, between the network size
scaling and the weight quantization, is more effective for hardware
optimization. For this study, the performances of fully-connected deep neural
networks (FCDNNs) and convolutional neural networks (CNNs) are evaluated while
changing the network complexity and the word-length of weights. Based on these
experiments, we present the effective compression ratio (ECR) to guide the
trade-off between the network size and the precision of weights when the
hardware resource is limited.
|
Conservative Contextual Linear Bandits | Safety is a desirable property that can immensely increase the applicability
of learning algorithms in real-world decision-making problems. It is much
easier for a company to deploy an algorithm that is safe, i.e., guaranteed to
perform at least as well as a baseline. In this paper, we study the issue of
safety in contextual linear bandits that have application in many different
fields including personalized ad recommendation in online marketing. We
formulate a notion of safety for this class of algorithms. We develop a safe
contextual linear bandit algorithm, called conservative linear UCB (CLUCB),
that simultaneously minimizes its regret and satisfies the safety constraint,
i.e., maintains its performance above a fixed percentage of the performance of
a baseline strategy, uniformly over time. We prove an upper-bound on the regret
of CLUCB and show that it can be decomposed into two terms: 1) an upper-bound
for the regret of the standard linear UCB algorithm that grows with the time
horizon and 2) a constant (does not grow with the time horizon) term that
accounts for the loss of being conservative in order to satisfy the safety
constraint. We empirically show that our algorithm is safe and validate our
theoretical analysis.
|
A Survey of Credit Card Fraud Detection Techniques: Data and Technique
Oriented Perspective | Credit card plays a very important rule in today's economy. It becomes an
unavoidable part of household, business and global activities. Although using
credit cards provides enormous benefits when used carefully and
responsibly,significant credit and financial damages may be caused by
fraudulent activities. Many techniques have been proposed to confront the
growth in credit card fraud. However, all of these techniques have the same
goal of avoiding the credit card fraud; each one has its own drawbacks,
advantages and characteristics. In this paper, after investigating difficulties
of credit card fraud detection, we seek to review the state of the art in
credit card fraud detection techniques, data sets and evaluation criteria.The
advantages and disadvantages of fraud detection methods are enumerated and
compared.Furthermore, a classification of mentioned techniques into two main
fraud detection approaches, namely, misuses (supervised) and anomaly detection
(unsupervised) is presented. Again, a classification of techniques is proposed
based on capability to process the numerical and categorical data sets.
Different data sets used in literature are then described and grouped into real
and synthesized data and the effective and common attributes are extracted for
further usage.Moreover, evaluation employed criterions in literature are
collected and discussed.Consequently, open issues for credit card fraud
detection are explained as guidelines for new researchers.
|
Pruning Convolutional Neural Networks for Resource Efficient Inference | We propose a new formulation for pruning convolutional kernels in neural
networks to enable efficient inference. We interleave greedy criteria-based
pruning with fine-tuning by backpropagation - a computationally efficient
procedure that maintains good generalization in the pruned network. We propose
a new criterion based on Taylor expansion that approximates the change in the
cost function induced by pruning network parameters. We focus on transfer
learning, where large pretrained networks are adapted to specialized tasks. The
proposed criterion demonstrates superior performance compared to other
criteria, e.g. the norm of kernel weights or feature map activation, for
pruning large CNNs after adaptation to fine-grained classification tasks
(Birds-200 and Flowers-102) relaying only on the first order gradient
information. We also show that pruning can lead to more than 10x theoretical
(5x practical) reduction in adapted 3D-convolutional filters with a small drop
in accuracy in a recurrent gesture classifier. Finally, we show results for the
large-scale ImageNet dataset to emphasize the flexibility of our approach.
|
Fast Video Classification via Adaptive Cascading of Deep Models | Recent advances have enabled "oracle" classifiers that can classify across
many classes and input distributions with high accuracy without retraining.
However, these classifiers are relatively heavyweight, so that applying them to
classify video is costly. We show that day-to-day video exhibits highly skewed
class distributions over the short term, and that these distributions can be
classified by much simpler models. We formulate the problem of detecting the
short-term skews online and exploiting models based on it as a new sequential
decision making problem dubbed the Online Bandit Problem, and present a new
algorithm to solve it. When applied to recognizing faces in TV shows and
movies, we realize end-to-end classification speedups of 2.4-7.8x/2.6-11.2x (on
GPU/CPU) relative to a state-of-the-art convolutional neural network, at
competitive accuracy.
|
Time Series Classification from Scratch with Deep Neural Networks: A
Strong Baseline | We propose a simple but strong baseline for time series classification from
scratch with deep neural networks. Our proposed baseline models are pure
end-to-end without any heavy preprocessing on the raw data or feature crafting.
The proposed Fully Convolutional Network (FCN) achieves premium performance to
other state-of-the-art approaches and our exploration of the very deep neural
networks with the ResNet structure is also competitive. The global average
pooling in our convolutional model enables the exploitation of the Class
Activation Map (CAM) to find out the contributing region in the raw data for
the specific labels. Our models provides a simple choice for the real world
application and a good starting point for the future research. An overall
analysis is provided to discuss the generalization capability of our models,
learned features, network structures and the classification semantics.
|
Dealing with Range Anxiety in Mean Estimation via Statistical Queries | We give algorithms for estimating the expectation of a given real-valued
function $\phi:X\to {\bf R}$ on a sample drawn randomly from some unknown
distribution $D$ over domain $X$, namely ${\bf E}_{{\bf x}\sim D}[\phi({\bf
x})]$. Our algorithms work in two well-studied models of restricted access to
data samples. The first one is the statistical query (SQ) model in which an
algorithm has access to an SQ oracle for the input distribution $D$ over $X$
instead of i.i.d. samples from $D$. Given a query function $\phi:X \to [0,1]$,
the oracle returns an estimate of ${\bf E}_{{\bf x}\sim D}[\phi({\bf x})]$
within some tolerance $\tau$. The second, is a model in which only a single bit
is communicated from each sample. In both of these models the error obtained
using a naive implementation would scale polynomially with the range of the
random variable $\phi({\bf x})$ (which might even be infinite). In contrast,
without restrictions on access to data the expected error scales with the
standard deviation of $\phi({\bf x})$. Here we give a simple algorithm whose
error scales linearly in standard deviation of $\phi({\bf x})$ and
logarithmically with an upper bound on the second moment of $\phi({\bf x})$.
As corollaries, we obtain algorithms for high dimensional mean estimation and
stochastic convex optimization in these models that work in more general
settings than previously known solutions.
|
Prototypical Recurrent Unit | Despite the great successes of deep learning, the effectiveness of deep
neural networks has not been understood at any theoretical depth. This work is
motivated by the thrust of developing a deeper understanding of recurrent
neural networks, particularly LSTM/GRU-like networks. As the highly complex
structure of the recurrent unit in LSTM and GRU networks makes them difficult
to analyze, our methodology in this research theme is to construct an
alternative recurrent unit that is as simple as possible and yet also captures
the key components of LSTM/GRU recurrent units. Such a unit can then be used
for the study of recurrent networks and its structural simplicity may allow
easier analysis. Towards that goal, we take a system-theoretic perspective to
design a new recurrent unit, which we call the prototypical recurrent unit
(PRU). Not only having minimal complexity, PRU is demonstrated experimentally
to have comparable performance to GRU and LSTM unit. This establishes PRU
networks as a prototype for future study of LSTM/GRU-like recurrent networks.
This paper also studies the memorization abilities of LSTM, GRU and PRU
networks, motivated by the folk belief that such networks possess long-term
memory. For this purpose, we design a simple and controllable task, called
``memorization problem'', where the networks are trained to memorize certain
targeted information. We show that the memorization performance of all three
networks depends on the amount of targeted information, the amount of
``interfering" information, and the state space dimension of the recurrent
unit. Experiments are also performed for another controllable task, the adding
problem, and similar conclusions are obtained.
|
Linear Thompson Sampling Revisited | We derive an alternative proof for the regret of Thompson sampling (\ts) in
the stochastic linear bandit setting. While we obtain a regret bound of order
$\widetilde{O}(d^{3/2}\sqrt{T})$ as in previous results, the proof sheds new
light on the functioning of the \ts. We leverage on the structure of the
problem to show how the regret is related to the sensitivity (i.e., the
gradient) of the objective function and how selecting optimal arms associated
to \textit{optimistic} parameters does control it. Thus we show that \ts can be
seen as a generic randomized algorithm where the sampling distribution is
designed to have a fixed probability of being optimistic, at the cost of an
additional $\sqrt{d}$ regret factor compared to a UCB-like approach.
Furthermore, we show that our proof can be readily applied to regularized
linear optimization and generalized linear model problems.
|
Efficient Stochastic Inference of Bitwise Deep Neural Networks | Recently published methods enable training of bitwise neural networks which
allow reduced representation of down to a single bit per weight. We present a
method that exploits ensemble decisions based on multiple stochastically
sampled network models to increase performance figures of bitwise neural
networks in terms of classification accuracy at inference. Our experiments with
the CIFAR-10 and GTSRB datasets show that the performance of such network
ensembles surpasses the performance of the high-precision base model. With this
technique we achieve 5.81% best classification error on CIFAR-10 test set using
bitwise networks. Concerning inference on embedded systems we evaluate these
bitwise networks using a hardware efficient stochastic rounding procedure. Our
work contributes to efficient embedded bitwise neural networks.
|
Variational Boosting: Iteratively Refining Posterior Approximations | We propose a black-box variational inference method to approximate
intractable distributions with an increasingly rich approximating class. Our
method, termed variational boosting, iteratively refines an existing
variational approximation by solving a sequence of optimization problems,
allowing the practitioner to trade computation time for accuracy. We show how
to expand the variational approximating class by incorporating additional
covariance structure and by introducing new components to form a mixture. We
apply variational boosting to synthetic and real statistical models, and show
that resulting posterior inferences compare favorably to existing posterior
approximation algorithms in both accuracy and efficiency.
|
Temporal Generative Adversarial Nets with Singular Value Clipping | In this paper, we propose a generative model, Temporal Generative Adversarial
Nets (TGAN), which can learn a semantic representation of unlabeled videos, and
is capable of generating videos. Unlike existing Generative Adversarial Nets
(GAN)-based methods that generate videos with a single generator consisting of
3D deconvolutional layers, our model exploits two different types of
generators: a temporal generator and an image generator. The temporal generator
takes a single latent variable as input and outputs a set of latent variables,
each of which corresponds to an image frame in a video. The image generator
transforms a set of such latent variables into a video. To deal with
instability in training of GAN with such advanced networks, we adopt a recently
proposed model, Wasserstein GAN, and propose a novel method to train it stably
in an end-to-end manner. The experimental results demonstrate the effectiveness
of our methods.
|
Deep Learning for the Classification of Lung Nodules | Deep learning, as a promising new area of machine learning, has attracted a
rapidly increasing attention in the field of medical imaging. Compared to the
conventional machine learning methods, deep learning requires no hand-tuned
feature extractor, and has shown a superior performance in many visual object
recognition applications. In this study, we develop a deep convolutional neural
network (CNN) and apply it to thoracic CT images for the classification of lung
nodules. We present the CNN architecture and classification accuracy for the
original images of lung nodules. In order to understand the features of lung
nodules, we further construct new datasets, based on the combination of
artificial geometric nodules and some transformations of the original images,
as well as a stochastic nodule shape model. It is found that simplistic
geometric nodules cannot capture the important features of lung nodules.
|
Scalable Adaptive Stochastic Optimization Using Random Projections | Adaptive stochastic gradient methods such as AdaGrad have gained popularity
in particular for training deep neural networks. The most commonly used and
studied variant maintains a diagonal matrix approximation to second order
information by accumulating past gradients which are used to tune the step size
adaptively. In certain situations the full-matrix variant of AdaGrad is
expected to attain better performance, however in high dimensions it is
computationally impractical. We present Ada-LR and RadaGrad two computationally
efficient approximations to full-matrix AdaGrad based on randomized
dimensionality reduction. They are able to capture dependencies between
features and achieve similar performance to full-matrix AdaGrad but at a much
smaller computational cost. We show that the regret of Ada-LR is close to the
regret of full-matrix AdaGrad which can have an up-to exponentially smaller
dependence on the dimension than the diagonal variant. Empirically, we show
that Ada-LR and RadaGrad perform similarly to full-matrix AdaGrad. On the task
of training convolutional neural networks as well as recurrent neural networks,
RadaGrad achieves faster convergence than diagonal AdaGrad.
|
Error analysis of regularized least-square regression with Fredholm
kernel | Learning with Fredholm kernel has attracted increasing attention recently
since it can effectively utilize the data information to improve the prediction
performance. Despite rapid progress on theoretical and experimental
evaluations, its generalization analysis has not been explored in learning
theory literature. In this paper, we establish the generalization bound of
least square regularized regression with Fredholm kernel, which implies that
the fast learning rate O(l^{-1}) can be reached under mild capacity conditions.
Simulated examples show that this Fredholm regression algorithm can achieve the
satisfactory prediction performance.
|
Probabilistic Duality for Parallel Gibbs Sampling without Graph Coloring | We present a new notion of probabilistic duality for random variables
involving mixture distributions. Using this notion, we show how to implement a
highly-parallelizable Gibbs sampler for weakly coupled discrete pairwise
graphical models with strictly positive factors that requires almost no
preprocessing and is easy to implement. Moreover, we show how our method can be
combined with blocking to improve mixing. Even though our method leads to
inferior mixing times compared to a sequential Gibbs sampler, we argue that our
method is still very useful for large dynamic networks, where factors are added
and removed on a continuous basis, as it is hard to maintain a graph coloring
in this setup. Similarly, our method is useful for parallelizing Gibbs sampling
in graphical models that do not allow for graph colorings with a small number
of colors such as densely connected graphs.
|
Training Sparse Neural Networks | Deep neural networks with lots of parameters are typically used for
large-scale computer vision tasks such as image classification. This is a
result of using dense matrix multiplications and convolutions. However, sparse
computations are known to be much more efficient. In this work, we train and
build neural networks which implicitly use sparse computations. We introduce
additional gate variables to perform parameter selection and show that this is
equivalent to using a spike-and-slab prior. We experimentally validate our
method on both small and large networks and achieve state-of-the-art
compression results for sparse neural network models.
|
On the convergence of gradient-like flows with noisy gradient input | In view of solving convex optimization problems with noisy gradient input, we
analyze the asymptotic behavior of gradient-like flows under stochastic
disturbances. Specifically, we focus on the widely studied class of mirror
descent schemes for convex programs with compact feasible regions, and we
examine the dynamics' convergence and concentration properties in the presence
of noise. In the vanishing noise limit, we show that the dynamics converge to
the solution set of the underlying problem (a.s.). Otherwise, when the noise is
persistent, we show that the dynamics are concentrated around interior
solutions in the long run, and they converge to boundary solutions that are
sufficiently "sharp". Finally, we show that a suitably rectified variant of the
method converges irrespective of the magnitude of the noise (or the structure
of the underlying convex program), and we derive an explicit estimate for its
rate of convergence.
|
Emergence of Compositional Representations in Restricted Boltzmann
Machines | Extracting automatically the complex set of features composing real
high-dimensional data is crucial for achieving high performance in
machine--learning tasks. Restricted Boltzmann Machines (RBM) are empirically
known to be efficient for this purpose, and to be able to generate distributed
and graded representations of the data. We characterize the structural
conditions (sparsity of the weights, low effective temperature, nonlinearities
in the activation functions of hidden units, and adaptation of fields
maintaining the activity in the visible layer) allowing RBM to operate in such
a compositional phase. Evidence is provided by the replica analysis of an
adequate statistical ensemble of random RBMs and by RBM trained on the
handwritten digits dataset MNIST.
|
Effective Deterministic Initialization for $k$-Means-Like Methods via
Local Density Peaks Searching | The $k$-means clustering algorithm is popular but has the following main
drawbacks: 1) the number of clusters, $k$, needs to be provided by the user in
advance, 2) it can easily reach local minima with randomly selected initial
centers, 3) it is sensitive to outliers, and 4) it can only deal with well
separated hyperspherical clusters. In this paper, we propose a Local Density
Peaks Searching (LDPS) initialization framework to address these issues. The
LDPS framework includes two basic components: one of them is the local density
that characterizes the density distribution of a data set, and the other is the
local distinctiveness index (LDI) which we introduce to characterize how
distinctive a data point is compared with its neighbors. Based on these two
components, we search for the local density peaks which are characterized with
high local densities and high LDIs to deal with 1) and 2). Moreover, we detect
outliers characterized with low local densities but high LDIs, and exclude them
out before clustering begins. Finally, we apply the LDPS initialization
framework to $k$-medoids, which is a variant of $k$-means and chooses data
samples as centers, with diverse similarity measures other than the Euclidean
distance to fix the last drawback of $k$-means. Combining the LDPS
initialization framework with $k$-means and $k$-medoids, we obtain two novel
clustering methods called LDPS-means and LDPS-medoids, respectively.
Experiments on synthetic data sets verify the effectiveness of the proposed
methods, especially when the ground truth of the cluster number $k$ is large.
Further, experiments on several real world data sets, Handwritten Pendigits,
Coil-20, Coil-100 and Olivetti Face Database, illustrate that our methods give
a superior performance than the analogous approaches on both estimating $k$ and
unsupervised object categorization.
|
Generalized Dropout | Deep Neural Networks often require good regularizers to generalize well.
Dropout is one such regularizer that is widely used among Deep Learning
practitioners. Recent work has shown that Dropout can also be viewed as
performing Approximate Bayesian Inference over the network parameters. In this
work, we generalize this notion and introduce a rich family of regularizers
which we call Generalized Dropout. One set of methods in this family, called
Dropout++, is a version of Dropout with trainable parameters. Classical Dropout
emerges as a special case of this method. Another member of this family selects
the width of neural network layers. Experiments show that these methods help in
improving generalization performance over Dropout.
|
Options Discovery with Budgeted Reinforcement Learning | We consider the problem of learning hierarchical policies for Reinforcement
Learning able to discover options, an option corresponding to a sub-policy over
a set of primitive actions. Different models have been proposed during the last
decade that usually rely on a predefined set of options. We specifically
address the problem of automatically discovering options in decision processes.
We describe a new learning model called Budgeted Option Neural Network (BONN)
able to discover options based on a budgeted learning objective. The BONN model
is evaluated on different classical RL problems, demonstrating both
quantitative and qualitative interesting results.
|
Probabilistic structure discovery in time series data | Existing methods for structure discovery in time series data construct
interpretable, compositional kernels for Gaussian process regression models.
While the learned Gaussian process model provides posterior mean and variance
estimates, typically the structure is learned via a greedy optimization
procedure. This restricts the space of possible solutions and leads to
over-confident uncertainty estimates. We introduce a fully Bayesian approach,
inferring a full posterior over structures, which more reliably captures the
uncertainty of the model.
|
Learning From Graph Neighborhoods Using LSTMs | Many prediction problems can be phrased as inferences over local
neighborhoods of graphs. The graph represents the interaction between entities,
and the neighborhood of each entity contains information that allows the
inferences or predictions. We present an approach for applying machine learning
directly to such graph neighborhoods, yielding predicitons for graph nodes on
the basis of the structure of their local neighborhood and the features of the
nodes in it. Our approach allows predictions to be learned directly from
examples, bypassing the step of creating and tuning an inference model or
summarizing the neighborhoods via a fixed set of hand-crafted features. The
approach is based on a multi-level architecture built from Long Short-Term
Memory neural nets (LSTMs); the LSTMs learn how to summarize the neighborhood
from data. We demonstrate the effectiveness of the proposed technique on a
synthetic example and on real-world data related to crowdsourced grading,
Bitcoin transactions, and Wikipedia edit reversions.
|
Unsupervised Learning for Lexicon-Based Classification | In lexicon-based classification, documents are assigned labels by comparing
the number of words that appear from two opposed lexicons, such as positive and
negative sentiment. Creating such words lists is often easier than labeling
instances, and they can be debugged by non-experts if classification
performance is unsatisfactory. However, there is little analysis or
justification of this classification heuristic. This paper describes a set of
assumptions that can be used to derive a probabilistic justification for
lexicon-based classification, as well as an analysis of its expected accuracy.
One key assumption behind lexicon-based classification is that all words in
each lexicon are equally predictive. This is rarely true in practice, which is
why lexicon-based approaches are usually outperformed by supervised classifiers
that learn distinct weights on each word from labeled instances. This paper
shows that it is possible to learn such weights without labeled data, by
leveraging co-occurrence statistics across the lexicons. This offers the best
of both worlds: light supervision in the form of lexicons, and data-driven
classification with higher accuracy than traditional word-counting heuristics.
|
Statistical Learning for OCR Text Correction | The accuracy of Optical Character Recognition (OCR) is crucial to the success
of subsequent applications used in text analyzing pipeline. Recent models of
OCR post-processing significantly improve the quality of OCR-generated text,
but are still prone to suggest correction candidates from limited observations
while insufficiently accounting for the characteristics of OCR errors. In this
paper, we show how to enlarge candidate suggestion space by using external
corpus and integrating OCR-specific features in a regression approach to
correct OCR-generated errors. The evaluation results show that our model can
correct 61.5% of the OCR-errors (considering the top 1 suggestion) and 71.5% of
the OCR-errors (considering the top 3 suggestions), for cases where the
theoretical correction upper-bound is 78%.
|
Associative Adversarial Networks | We propose a higher-level associative memory for learning adversarial
networks. Generative adversarial network (GAN) framework has a discriminator
and a generator network. The generator (G) maps white noise (z) to data samples
while the discriminator (D) maps data samples to a single scalar. To do so, G
learns how to map from high-level representation space to data space, and D
learns to do the opposite. We argue that higher-level representation spaces
need not necessarily follow a uniform probability distribution. In this work,
we use Restricted Boltzmann Machines (RBMs) as a higher-level associative
memory and learn the probability distribution for the high-level features
generated by D. The associative memory samples its underlying probability
distribution and G learns how to map these samples to data space. The proposed
associative adversarial networks (AANs) are generative models in the
higher-levels of the learning, and use adversarial non-stochastic models D and
G for learning the mapping between data and higher-level representation spaces.
Experiments show the potential of the proposed networks.
|
Measuring Sample Quality with Diffusions | Stein's method for measuring convergence to a continuous target distribution
relies on an operator characterizing the target and Stein factor bounds on the
solutions of an associated differential equation. While such operators and
bounds are readily available for a diversity of univariate targets, few
multivariate targets have been analyzed. We introduce a new class of
characterizing operators based on Ito diffusions and develop explicit
multivariate Stein factor bounds for any target with a fast-coupling Ito
diffusion. As example applications, we develop computable and
convergence-determining diffusion Stein discrepancies for log-concave,
heavy-tailed, and multimodal targets and use these quality measures to select
the hyperparameters of biased Markov chain Monte Carlo (MCMC) samplers, compare
random and deterministic quadrature rules, and quantify bias-variance tradeoffs
in approximate MCMC. Our results establish a near-linear relationship between
diffusion Stein discrepancies and Wasserstein distances, improving upon past
work even for strongly log-concave targets. The exposed relationship between
Stein factors and Markov process coupling may be of independent interest.
|
Robust end-to-end deep audiovisual speech recognition | Speech is one of the most effective ways of communication among humans. Even
though audio is the most common way of transmitting speech, very important
information can be found in other modalities, such as vision. Vision is
particularly useful when the acoustic signal is corrupted. Multi-modal speech
recognition however has not yet found wide-spread use, mostly because the
temporal alignment and fusion of the different information sources is
challenging.
This paper presents an end-to-end audiovisual speech recognizer (AVSR), based
on recurrent neural networks (RNN) with a connectionist temporal classification
(CTC) loss function. CTC creates sparse "peaky" output activations, and we
analyze the differences in the alignments of output targets (phonemes or
visemes) between audio-only, video-only, and audio-visual feature
representations. We present the first such experiments on the large vocabulary
IBM ViaVoice database, which outperform previously published approaches on
phone accuracy in clean and noisy conditions.
|
Spatial contrasting for deep unsupervised learning | Convolutional networks have marked their place over the last few years as the
best performing model for various visual tasks. They are, however, most suited
for supervised learning from large amounts of labeled data. Previous attempts
have been made to use unlabeled data to improve model performance by applying
unsupervised techniques. These attempts require different architectures and
training methods. In this work we present a novel approach for unsupervised
training of Convolutional networks that is based on contrasting between spatial
regions within images. This criterion can be employed within conventional
neural networks and trained using standard techniques such as SGD and
back-propagation, thus complementing supervised methods.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.