title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Online Machine Learning Techniques for Predicting Operator Performance | This thesis explores a number of online machine learning algorithms. From a
theoret- ical perspective, it assesses their employability for a particular
function approximation problem where the analytical models fall short.
Furthermore, it discusses the applica- tion of theoretically suitable learning
algorithms to the function approximation problem at hand through an efficient
implementation that exploits various computational and mathematical shortcuts.
Finally, this thesis work evaluates the implemented learning algorithms
according to various evaluation criteria through rigorous testing.
|
Do logarithmic proximity measures outperform plain ones in graph
clustering? | We consider a number of graph kernels and proximity measures including
commute time kernel, regularized Laplacian kernel, heat kernel, exponential
diffusion kernel (also called "communicability"), etc., and the corresponding
distances as applied to clustering nodes in random graphs and several
well-known datasets. The model of generating random graphs involves edge
probabilities for the pairs of nodes that belong to the same class or different
predefined classes of nodes. It turns out that in most cases, logarithmic
measures (i.e., measures resulting after taking logarithm of the proximities)
perform better while distinguishing underlying classes than the "plain"
measures. A comparison in terms of reject curves of inter-class and intra-class
distances confirms this conclusion. A similar conclusion can be made for
several well-known datasets. A possible origin of this effect is that most
kernels have a multiplicative nature, while the nature of distances used in
cluster algorithms is an additive one (cf. the triangle inequality). The
logarithmic transformation is a tool to transform the first nature to the
second one. Moreover, some distances corresponding to the logarithmic measures
possess a meaningful cutpoint additivity property. In our experiments, the
leader is usually the logarithmic Communicability measure. However, we indicate
some more complicated cases in which other measures, typically, Communicability
and plain Walk, can be the winners.
|
Decentralized Dynamic Discriminative Dictionary Learning | We consider discriminative dictionary learning in a distributed online
setting, where a network of agents aims to learn a common set of dictionary
elements of a feature space and model parameters while sequentially receiving
observations. We formulate this problem as a distributed stochastic program
with a non-convex objective and present a block variant of the Arrow-Hurwicz
saddle point algorithm to solve it. Using Lagrange multipliers to penalize the
discrepancy between them, only neighboring nodes exchange model information. We
show that decisions made with this saddle point algorithm asymptotically
achieve a first-order stationarity condition on average.
|
An evaluation of randomized machine learning methods for redundant data:
Predicting short and medium-term suicide risk from administrative records and
risk assessments | Accurate prediction of suicide risk in mental health patients remains an open
problem. Existing methods including clinician judgments have acceptable
sensitivity, but yield many false positives. Exploiting administrative data has
a great potential, but the data has high dimensionality and redundancies in the
recording processes. We investigate the efficacy of three most effective
randomized machine learning techniques random forests, gradient boosting
machines, and deep neural nets with dropout in predicting suicide risk. Using a
cohort of mental health patients from a regional Australian hospital, we
compare the predictive performance with popular traditional approaches
clinician judgments based on a checklist, sparse logistic regression and
decision trees. The randomized methods demonstrated robustness against data
redundancies and superior predictive performance on AUC and F-measure.
|
Deep Motif: Visualizing Genomic Sequence Classifications | This paper applies a deep convolutional/highway MLP framework to classify
genomic sequences on the transcription factor binding site task. To make the
model understandable, we propose an optimization driven strategy to extract
"motifs", or symbolic patterns which visualize the positive class learned by
the network. We show that our system, Deep Motif (DeMo), extracts motifs that
are similar to, and in some cases outperform the current well known motifs. In
addition, we find that a deeper model consisting of multiple convolutional and
highway layers can outperform a single convolutional and fully connected layer
in the previous state-of-the-art.
|
Linear Bandit algorithms using the Bootstrap | This study presents two new algorithms for solving linear stochastic bandit
problems. The proposed methods use an approach from non-parametric statistics
called bootstrapping to create confidence bounds. This is achieved without
making any assumptions about the distribution of noise in the underlying
system. We present the X-Random and X-Fixed bootstrap bandits which correspond
to the two well-known approaches for conducting bootstraps on models, in the
literature. The proposed methods are compared to other popular solutions for
linear stochastic bandit problems, namely, OFUL, LinUCB and Thompson Sampling.
The comparisons are carried out using a simulation study on a hierarchical
probability meta-model, built from published data of experiments, which are run
on real systems. The model representing the response surfaces is conceptualized
as a Bayesian Network which is presented with varying degrees of noise for the
simulations. One of the proposed methods, X-Random bootstrap, performs better
than the baselines in-terms of cumulative regret across various degrees of
noise and different number of trials. In certain settings the cumulative regret
of this method is less than half of the best baseline. The X-Fixed bootstrap
performs comparably in most situations and particularly well when the number of
trials is low. The study concludes that these algorithms could be a preferred
alternative for solving linear bandit problems, especially when the
distribution of the noise in the system is unknown.
|
A Bayesian Approach to Policy Recognition and State Representation
Learning | Learning from demonstration (LfD) is the process of building behavioral
models of a task from demonstrations provided by an expert. These models can be
used e.g. for system control by generalizing the expert demonstrations to
previously unencountered situations. Most LfD methods, however, make strong
assumptions about the expert behavior, e.g. they assume the existence of a
deterministic optimal ground truth policy or require direct monitoring of the
expert's controls, which limits their practical use as part of a general system
identification framework. In this work, we consider the LfD problem in a more
general setting where we allow for arbitrary stochastic expert policies,
without reasoning about the optimality of the demonstrations. Following a
Bayesian methodology, we model the full posterior distribution of possible
expert controllers that explain the provided demonstration data. Moreover, we
show that our methodology can be applied in a nonparametric context to infer
the complexity of the state representation used by the expert, and to learn
task-appropriate partitionings of the system state space.
|
Fast rates with high probability in exp-concave statistical learning | We present an algorithm for the statistical learning setting with a bounded
exp-concave loss in $d$ dimensions that obtains excess risk $O(d
\log(1/\delta)/n)$ with probability at least $1 - \delta$. The core technique
is to boost the confidence of recent in-expectation $O(d/n)$ excess risk bounds
for empirical risk minimization (ERM), without sacrificing the rate, by
leveraging a Bernstein condition which holds due to exp-concavity. We also show
that with probability $1 - \delta$ the standard ERM method obtains excess risk
$O(d (\log(n) + \log(1/\delta))/n)$. We further show that a regret bound for
any online learner in this setting translates to a high probability excess risk
bound for the corresponding online-to-batch conversion of the online learner.
Lastly, we present two high probability bounds for the exp-concave model
selection aggregation problem that are quantile-adaptive in a certain sense.
The first bound is a purely exponential weights type algorithm, obtains a
nearly optimal rate, and has no explicit dependence on the Lipschitz continuity
of the loss. The second bound requires Lipschitz continuity but obtains the
optimal rate.
|
Single Channel Speech Enhancement Using Outlier Detection | Distortion of the underlying speech is a common problem for single-channel
speech enhancement algorithms, and hinders such methods from being used more
extensively. A dictionary based speech enhancement method that emphasizes
preserving the underlying speech is proposed. Spectral patches of clean speech
are sampled and clustered to train a dictionary. Given a noisy speech spectral
patch, the best matching dictionary entry is selected and used to estimate the
noise power at each time-frequency bin. The noise estimation step is formulated
as an outlier detection problem, where the noise at each bin is assumed present
only if it is an outlier to the corresponding bin of the best matching
dictionary entry. This framework assigns higher priority in removing spectral
elements that strongly deviate from a typical spoken unit stored in the trained
dictionary. Even without the aid of a separate noise model, this method can
achieve significant noise reduction for various non-stationary noises, while
effectively preserving the underlying speech in more challenging noisy
environments.
|
Learning from the memory of Atari 2600 | We train a number of neural networks to play games Bowling, Breakout and
Seaquest using information stored in the memory of a video game console Atari
2600. We consider four models of neural networks which differ in size and
architecture: two networks which use only information contained in the RAM and
two mixed networks which use both information in the RAM and information from
the screen. As the benchmark we used the convolutional model proposed in NIPS
and received comparable results in all considered games. Quite surprisingly, in
the case of Seaquest we were able to train RAM-only agents which behave better
than the benchmark screen-only agent. Mixing screen and RAM did not lead to an
improved performance comparing to screen-only and RAM-only agents.
|
Accelerating Deep Learning with Shrinkage and Recall | Deep Learning is a very powerful machine learning model. Deep Learning trains
a large number of parameters for multiple layers and is very slow when data is
in large scale and the architecture size is large. Inspired from the shrinking
technique used in accelerating computation of Support Vector Machines (SVM)
algorithm and screening technique used in LASSO, we propose a shrinking Deep
Learning with recall (sDLr) approach to speed up deep learning computation. We
experiment shrinking Deep Learning with recall (sDLr) using Deep Neural Network
(DNN), Deep Belief Network (DBN) and Convolution Neural Network (CNN) on 4 data
sets. Results show that the speedup using shrinking Deep Learning with recall
(sDLr) can reach more than 2.0 while still giving competitive classification
performance.
|
Boltzmann meets Nash: Energy-efficient routing in optical networks under
uncertainty | Motivated by the massive deployment of power-hungry data centers for service
provisioning, we examine the problem of routing in optical networks with the
aim of minimizing traffic-driven power consumption. To tackle this issue,
routing must take into account energy efficiency as well as capacity
considerations; moreover, in rapidly-varying network environments, this must be
accomplished in a real-time, distributed manner that remains robust in the
presence of random disturbances and noise. In view of this, we derive a pricing
scheme whose Nash equilibria coincide with the network's socially optimum
states, and we propose a distributed learning method based on the Boltzmann
distribution of statistical mechanics. Using tools from stochastic calculus, we
show that the resulting Boltzmann routing scheme exhibits remarkable
convergence properties under uncertainty: specifically, the long-term average
of the network's power consumption converges within $\varepsilon$ of its
minimum value in time which is at most $\tilde O(1/\varepsilon^2)$,
irrespective of the fluctuations' magnitude; additionally, if the network
admits a strict, non-mixing optimum state, the algorithm converges to it -
again, no matter the noise level. Our analysis is supplemented by extensive
numerical simulations which show that Boltzmann routing can lead to a
significant decrease in power consumption over basic, shortest-path routing
schemes in realistic network conditions.
|
Classification of Human Whole-Body Motion using Hidden Markov Models | Human motion plays an important role in many fields. Large databases exist
that store and make available recordings of human motions. However, annotating
each motion with multiple labels is a cumbersome and error-prone process. This
bachelor's thesis presents different approaches to solve the multi-label
classification problem using Hidden Markov Models (HMMs). First, different
features that can be directly obtained from the raw data are introduced. Next,
additional features are derived to improve classification performance. These
features are then used to perform the multi-label classification using two
different approaches. The first approach simply transforms the multi-label
problem into a multi-class problem. The second, novel approach solves the same
problem without the need to construct a transformation by predicting the labels
directly from the likelihood scores. The second approach scales linearly with
the number of labels whereas the first approach is subject to combinatorial
explosion. All aspects of the classification process are evaluated on a data
set that consists of 454 motions. System 1 achieves an accuracy of 98.02% and
system 2 an accuracy of 93.39% on the test set.
|
On the Convergence of A Family of Robust Losses for Stochastic Gradient
Descent | The convergence of Stochastic Gradient Descent (SGD) using convex loss
functions has been widely studied. However, vanilla SGD methods using convex
losses cannot perform well with noisy labels, which adversely affect the update
of the primal variable in SGD methods. Unfortunately, noisy labels are
ubiquitous in real world applications such as crowdsourcing. To handle noisy
labels, in this paper, we present a family of robust losses for SGD methods. By
employing our robust losses, SGD methods successfully reduce negative effects
caused by noisy labels on each update of the primal variable. We not only
reveal that the convergence rate is O(1/T) for SGD methods using robust losses,
but also provide the robustness analysis on two representative robust losses.
Comprehensive experimental results on six real-world datasets show that SGD
methods using robust losses are obviously more robust than other baseline
methods in most situations with fast convergence.
|
Maximal Sparsity with Deep Networks? | The iterations of many sparse estimation algorithms are comprised of a fixed
linear filter cascaded with a thresholding nonlinearity, which collectively
resemble a typical neural network layer. Consequently, a lengthy sequence of
algorithm iterations can be viewed as a deep network with shared, hand-crafted
layer weights. It is therefore quite natural to examine the degree to which a
learned network model might act as a viable surrogate for traditional sparse
estimation in domains where ample training data is available. While the
possibility of a reduced computational budget is readily apparent when a
ceiling is imposed on the number of layers, our work primarily focuses on
estimation accuracy. In particular, it is well-known that when a signal
dictionary has coherent columns, as quantified by a large RIP constant, then
most tractable iterative algorithms are unable to find maximally sparse
representations. In contrast, we demonstrate both theoretically and empirically
the potential for a trained deep network to recover minimal $\ell_0$-norm
representations in regimes where existing methods fail. The resulting system is
deployed on a practical photometric stereo estimation problem, where the goal
is to remove sparse outliers that can disrupt the estimation of surface normals
from a 3D scene.
|
A Tight Bound of Hard Thresholding | This paper is concerned with the hard thresholding operator which sets all
but the $k$ largest absolute elements of a vector to zero. We establish a {\em
tight} bound to quantitatively characterize the deviation of the thresholded
solution from a given signal. Our theoretical result is universal in the sense
that it holds for all choices of parameters, and the underlying analysis
depends only on fundamental arguments in mathematical optimization. We discuss
the implications for two domains:
Compressed Sensing. On account of the crucial estimate, we bridge the
connection between the restricted isometry property (RIP) and the sparsity
parameter for a vast volume of hard thresholding based algorithms, which
renders an improvement on the RIP condition especially when the true sparsity
is unknown. This suggests that in essence, many more kinds of sensing matrices
or fewer measurements are admissible for the data acquisition procedure.
Machine Learning. In terms of large-scale machine learning, a significant yet
challenging problem is learning accurate sparse models in an efficient manner.
In stark contrast to prior work that attempted the $\ell_1$-relaxation for
promoting sparsity, we present a novel stochastic algorithm which performs hard
thresholding in each iteration, hence ensuring such parsimonious solutions.
Equipped with the developed bound, we prove the {\em global linear convergence}
for a number of prevalent statistical models under mild assumptions, even
though the problem turns out to be non-convex.
|
Copeland Dueling Bandit Problem: Regret Lower Bound, Optimal Algorithm,
and Computationally Efficient Algorithm | We study the K-armed dueling bandit problem, a variation of the standard
stochastic bandit problem where the feedback is limited to relative comparisons
of a pair of arms. The hardness of recommending Copeland winners, the arms that
beat the greatest number of other arms, is characterized by deriving an
asymptotic regret bound. We propose Copeland Winners Relative Minimum Empirical
Divergence (CW-RMED) and derive an asymptotically optimal regret bound for it.
However, it is not known whether the algorithm can be efficiently computed or
not. To address this issue, we devise an efficient version (ECW-RMED) and
derive its asymptotic regret bound. Experimental comparisons of dueling bandit
algorithms show that ECW-RMED significantly outperforms existing ones.
|
A note on adjusting $R^2$ for using with cross-validation | We show how to adjust the coefficient of determination ($R^2$) when used for
measuring predictive accuracy via leave-one-out cross-validation.
|
Not Just a Black Box: Learning Important Features Through Propagating
Activation Differences | Note: This paper describes an older version of DeepLIFT. See
https://arxiv.org/abs/1704.02685 for the newer version. Original abstract
follows: The purported "black box" nature of neural networks is a barrier to
adoption in applications where interpretability is essential. Here we present
DeepLIFT (Learning Important FeaTures), an efficient and effective method for
computing importance scores in a neural network. DeepLIFT compares the
activation of each neuron to its 'reference activation' and assigns
contribution scores according to the difference. We apply DeepLIFT to models
trained on natural images and genomic data, and show significant advantages
over gradient-based methods.
|
Rank Ordered Autoencoders | A new method for the unsupervised learning of sparse representations using
autoencoders is proposed and implemented by ordering the output of the hidden
units by their activation value and progressively reconstructing the input in
this order. This can be done efficiently in parallel with the use of cumulative
sums and sorting only slightly increasing the computational costs. Minimizing
the difference of this progressive reconstruction with respect to the input can
be seen as minimizing the number of active output units required for the
reconstruction of the input. The model thus learns to reconstruct optimally
using the least number of active output units. This leads to high sparsity
without the need for extra hyperparameters, the amount of sparsity is instead
implicitly learned by minimizing this progressive reconstruction error. Results
of the trained model are given for patches of the CIFAR10 dataset, showing
rapid convergence of features and extremely sparse output activations while
maintaining a minimal reconstruction error and showing extreme robustness to
overfitting. Additionally the reconstruction as function of number of active
units is presented which shows the autoencoder learns a rank order code over
the input where the highest ranked units correspond to the highest decrease in
reconstruction error.
|
DCTNet and PCANet for acoustic signal feature extraction | We introduce the use of DCTNet, an efficient approximation and alternative to
PCANet, for acoustic signal classification. In PCANet, the eigenfunctions of
the local sample covariance matrix (PCA) are used as filterbanks for
convolution and feature extraction. When the eigenfunctions are well
approximated by the Discrete Cosine Transform (DCT) functions, each layer of of
PCANet and DCTNet is essentially a time-frequency representation. We relate
DCTNet to spectral feature representation methods, such as the the short time
Fourier transform (STFT), spectrogram and linear frequency spectral
coefficients (LFSC). Experimental results on whale vocalization data show that
DCTNet improves classification rate, demonstrating DCTNet's applicability to
signal processing problems such as underwater acoustics.
|
Cross-Graph Learning of Multi-Relational Associations | Cross-graph Relational Learning (CGRL) refers to the problem of predicting
the strengths or labels of multi-relational tuples of heterogeneous object
types, through the joint inference over multiple graphs which specify the
internal connections among each type of objects. CGRL is an open challenge in
machine learning due to the daunting number of all possible tuples to deal with
when the numbers of nodes in multiple graphs are large, and because the labeled
training instances are extremely sparse as typical. Existing methods such as
tensor factorization or tensor-kernel machines do not work well because of the
lack of convex formulation for the optimization of CGRL models, the poor
scalability of the algorithms in handling combinatorial numbers of tuples,
and/or the non-transductive nature of the learning methods which limits their
ability to leverage unlabeled data in training. This paper proposes a novel
framework which formulates CGRL as a convex optimization problem, enables
transductive learning using both labeled and unlabeled tuples, and offers a
scalable algorithm that guarantees the optimal solution and enjoys a linear
time complexity with respect to the sizes of input graphs. In our experiments
with a subset of DBLP publication records and an Enzyme multi-source dataset,
the proposed method successfully scaled to the large cross-graph inference
problem, and outperformed other representative approaches significantly.
|
DeepPicker: a Deep Learning Approach for Fully Automated Particle
Picking in Cryo-EM | Particle picking is a time-consuming step in single-particle analysis and
often requires significant interventions from users, which has become a
bottleneck for future automated electron cryo-microscopy (cryo-EM). Here we
report a deep learning framework, called DeepPicker, to address this problem
and fill the current gaps toward a fully automated cryo-EM pipeline. DeepPicker
employs a novel cross-molecule training strategy to capture common features of
particles from previously-analyzed micrographs, and thus does not require any
human intervention during particle picking. Tests on the recently-published
cryo-EM data of three complexes have demonstrated that our deep learning based
scheme can successfully accomplish the human-level particle picking process and
identify a sufficient number of particles that are comparable to those manually
by human experts. These results indicate that DeepPicker can provide a
practically useful tool to significantly reduce the time and manual effort
spent in single-particle analysis and thus greatly facilitate high-resolution
cryo-EM structure determination.
|
Energy Disaggregation for Real-Time Building Flexibility Detection | Energy is a limited resource which has to be managed wisely, taking into
account both supply-demand matching and capacity constraints in the
distribution grid. One aspect of the smart energy management at the building
level is given by the problem of real-time detection of flexible demand
available. In this paper we propose the use of energy disaggregation techniques
to perform this task. Firstly, we investigate the use of existing
classification methods to perform energy disaggregation. A comparison is
performed between four classifiers, namely Naive Bayes, k-Nearest Neighbors,
Support Vector Machine and AdaBoost. Secondly, we propose the use of Restricted
Boltzmann Machine to automatically perform feature extraction. The extracted
features are then used as inputs to the four classifiers and consequently shown
to improve their accuracy. The efficiency of our approach is demonstrated on a
real database consisting of detailed appliance-level measurements with high
temporal resolution, which has been used for energy disaggregation in previous
studies, namely the REDD. The results show robustness and good generalization
capabilities to newly presented buildings with at least 96% accuracy.
|
Automatic LQR Tuning Based on Gaussian Process Global Optimization | This paper proposes an automatic controller tuning framework based on linear
optimal control combined with Bayesian optimization. With this framework, an
initial set of controller gains is automatically improved according to a
pre-defined performance objective evaluated from experimental data. The
underlying Bayesian optimization algorithm is Entropy Search, which represents
the latent objective as a Gaussian process and constructs an explicit belief
over the location of the objective minimum. This is used to maximize the
information gain from each experimental evaluation. Thus, this framework shall
yield improved controllers with fewer evaluations compared to alternative
approaches. A seven-degree-of-freedom robot arm balancing an inverted pole is
used as the experimental demonstrator. Results of a two- and four-dimensional
tuning problems highlight the method's potential for automatic controller
tuning on robotic platforms.
|
Training Neural Networks Without Gradients: A Scalable ADMM Approach | With the growing importance of large network models and enormous training
datasets, GPUs have become increasingly necessary to train neural networks.
This is largely because conventional optimization algorithms rely on stochastic
gradient methods that don't scale well to large numbers of cores in a cluster
setting. Furthermore, the convergence of all gradient methods, including batch
methods, suffers from common problems like saturation effects, poor
conditioning, and saddle points. This paper explores an unconventional training
method that uses alternating direction methods and Bregman iteration to train
networks without gradient descent steps. The proposed method reduces the
network training problem to a sequence of minimization sub-steps that can each
be solved globally in closed form. The proposed method is advantageous because
it avoids many of the caveats that make gradient methods slow on highly
non-convex problems. The method exhibits strong scaling in the distributed
setting, yielding linear speedups even when split over thousands of cores.
|
Low-Complexity Stochastic Generalized Belief Propagation | The generalized belief propagation (GBP), introduced by Yedidia et al., is an
extension of the belief propagation (BP) algorithm, which is widely used in
different problems involved in calculating exact or approximate marginals of
probability distributions. In many problems, it has been observed that the
accuracy of GBP considerably outperforms that of BP. However, because in
general the computational complexity of GBP is higher than BP, its application
is limited in practice.
In this paper, we introduce a stochastic version of GBP called stochastic
generalized belief propagation (SGBP) that can be considered as an extension to
the stochastic BP (SBP) algorithm introduced by Noorshams et al. They have
shown that SBP reduces the complexity per iteration of BP by an order of
magnitude in alphabet size. In contrast to SBP, SGBP can reduce the computation
complexity if certain topological conditions are met by the region graph
associated to a graphical model. However, this reduction can be larger than
only one order of magnitude in alphabet size. In this paper, we characterize
these conditions and the amount of computation gain that we can obtain by using
SGBP. Finally, using similar proof techniques employed by Noorshams et al., for
general graphical models satisfy contraction conditions, we prove the
asymptotic convergence of SGBP to the unique GBP fixed point, as well as
providing non-asymptotic upper bounds on the mean square error and on the high
probability error.
|
Concentrated Differential Privacy: Simplifications, Extensions, and
Lower Bounds | "Concentrated differential privacy" was recently introduced by Dwork and
Rothblum as a relaxation of differential privacy, which permits sharper
analyses of many privacy-preserving computations. We present an alternative
formulation of the concept of concentrated differential privacy in terms of the
Renyi divergence between the distributions obtained by running an algorithm on
neighboring inputs. With this reformulation in hand, we prove sharper
quantitative results, establish lower bounds, and raise a few new questions. We
also unify this approach with approximate differential privacy by giving an
appropriate definition of "approximate concentrated differential privacy."
|
Function-Specific Mixing Times and Concentration Away from Equilibrium | Slow mixing is the central hurdle when working with Markov chains, especially
those used for Monte Carlo approximations (MCMC). In many applications, it is
only of interest to estimate the stationary expectations of a small set of
functions, and so the usual definition of mixing based on total variation
convergence may be too conservative. Accordingly, we introduce
function-specific analogs of mixing times and spectral gaps, and use them to
prove Hoeffding-like function-specific concentration inequalities. These
results show that it is possible for empirical expectations of functions to
concentrate long before the underlying chain has mixed in the classical sense,
and we show that the concentration rates we achieve are optimal up to
constants. We use our techniques to derive confidence intervals that are
sharper than those implied by both classical Markov chain Hoeffding bounds and
Berry-Esseen-corrected CLT bounds. For applications that require testing,
rather than point estimation, we show similar improvements over recent
sequential testing results for MCMC. We conclude by applying our framework to
real data examples of MCMC, providing evidence that our theory is both accurate
and relevant to practice.
|
ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement
Learning | The recent advances in deep neural networks have led to effective
vision-based reinforcement learning methods that have been employed to obtain
human-level controllers in Atari 2600 games from pixel data. Atari 2600 games,
however, do not resemble real-world tasks since they involve non-realistic 2D
environments and the third-person perspective. Here, we propose a novel
test-bed platform for reinforcement learning research from raw visual
information which employs the first-person perspective in a semi-realistic 3D
world. The software, called ViZDoom, is based on the classical first-person
shooter video game, Doom. It allows developing bots that play the game using
the screen buffer. ViZDoom is lightweight, fast, and highly customizable via a
convenient mechanism of user scenarios. In the experimental part, we test the
environment by trying to learn bots for two scenarios: a basic move-and-shoot
task and a more complex maze-navigation problem. Using convolutional deep
neural networks with Q-learning and experience replay, for both scenarios, we
were able to train competent bots, which exhibit human-like behaviors. The
results confirm the utility of ViZDoom as an AI research platform and imply
that visual reinforcement learning in 3D realistic first-person perspective
environments is feasible.
|
Some Simulation Results for Emphatic Temporal-Difference Learning
Algorithms | This is a companion note to our recent study of the weak convergence
properties of constrained emphatic temporal-difference learning (ETD)
algorithms from a theoretic perspective. It supplements the latter analysis
with simulation results and illustrates the behavior of some of the ETD
algorithms using three example problems.
|
Distributed Learning with Infinitely Many Hypotheses | We consider a distributed learning setup where a network of agents
sequentially access realizations of a set of random variables with unknown
distributions. The network objective is to find a parametrized distribution
that best describes their joint observations in the sense of the
Kullback-Leibler divergence. Apart from recent efforts in the literature, we
analyze the case of countably many hypotheses and the case of a continuum of
hypotheses. We provide non-asymptotic bounds for the concentration rate of the
agents' beliefs around the correct hypothesis in terms of the number of agents,
the network parameters, and the learning abilities of the agents. Additionally,
we provide a novel motivation for a general set of distributed Non-Bayesian
update rules as instances of the distributed stochastic mirror descent
algorithm.
|
Adobe-MIT submission to the DSTC 4 Spoken Language Understanding pilot
task | The Dialog State Tracking Challenge 4 (DSTC 4) proposes several pilot tasks.
In this paper, we focus on the spoken language understanding pilot task, which
consists of tagging a given utterance with speech acts and semantic slots. We
compare different classifiers: the best system obtains 0.52 and 0.67 F1-scores
on the test set for speech act recognition for the tourist and the guide
respectively, and 0.52 F1-score for semantic tagging for both the guide and the
tourist.
|
Robust Dialog State Tracking for Large Ontologies | The Dialog State Tracking Challenge 4 (DSTC 4) differentiates itself from the
previous three editions as follows: the number of slot-value pairs present in
the ontology is much larger, no spoken language understanding output is given,
and utterances are labeled at the subdialog level. This paper describes a novel
dialog state tracking method designed to work robustly under these conditions,
using elaborate string matching, coreference resolution tailored for dialogs
and a few other improvements. The method can correctly identify many values
that are not explicitly present in the utterance. On the final evaluation, our
method came in first among 7 competing teams and 24 entries. The F1-score
achieved by our method was 9 and 7 percentage points higher than that of the
runner-up for the utterance-level evaluation and for the subdialog-level
evaluation, respectively.
|
All Weather Perception: Joint Data Association, Tracking, and
Classification for Autonomous Ground Vehicles | A novel probabilistic perception algorithm is presented as a real-time joint
solution to data association, object tracking, and object classification for an
autonomous ground vehicle in all-weather conditions. The presented algorithm
extends a Rao-Blackwellized Particle Filter originally built with a particle
filter for data association and a Kalman filter for multi-object tracking
(Miller et al. 2011a) to now also include multiple model tracking for
classification. Additionally a state-of-the-art vision detection algorithm that
includes heading information for autonomous ground vehicle (AGV) applications
was implemented. Cornell's AGV from the DARPA Urban Challenge was upgraded and
used to experimentally examine if and how state-of-the-art vision algorithms
can complement or replace lidar and radar sensors. Sensor and algorithm
performance in adverse weather and lighting conditions is tested. Experimental
evaluation demonstrates robust all-weather data association, tracking, and
classification where camera, lidar, and radar sensors complement each other
inside the joint probabilistic perception algorithm.
|
Distributed stochastic optimization for deep learning (thesis) | We study the problem of how to distribute the training of large-scale deep
learning models in the parallel computing environment. We propose a new
distributed stochastic optimization method called Elastic Averaging SGD
(EASGD). We analyze the convergence rate of the EASGD method in the synchronous
scenario and compare its stability condition with the existing ADMM method in
the round-robin scheme. An asynchronous and momentum variant of the EASGD
method is applied to train deep convolutional neural networks for image
classification on the CIFAR and ImageNet datasets. Our approach accelerates the
training and furthermore achieves better test accuracy. It also requires a much
smaller amount of communication than other common baseline approaches such as
the DOWNPOUR method.
We then investigate the limit in speedup of the initial and the asymptotic
phase of the mini-batch SGD, the momentum SGD, and the EASGD methods. We find
that the spread of the input data distribution has a big impact on their
initial convergence rate and stability region. We also find a surprising
connection between the momentum SGD and the EASGD method with a negative moving
average rate. A non-convex case is also studied to understand when EASGD can
get trapped by a saddle point.
Finally, we scale up the EASGD method by using a tree structured network
topology. We show empirically its advantage and challenge. We also establish a
connection between the EASGD and the DOWNPOUR method with the classical Jacobi
and the Gauss-Seidel method, thus unifying a class of distributed stochastic
optimization methods.
|
Neural Autoregressive Distribution Estimation | We present Neural Autoregressive Distribution Estimation (NADE) models, which
are neural network architectures applied to the problem of unsupervised
distribution and density estimation. They leverage the probability product rule
and a weight sharing scheme inspired from restricted Boltzmann machines, to
yield an estimator that is both tractable and has good generalization
performance. We discuss how they achieve competitive performance in modeling
both binary and real-valued observations. We also present how deep NADE models
can be trained to be agnostic to the ordering of input dimensions used by the
autoregressive product rule decomposition. Finally, we also show how to exploit
the topological structure of pixels in images using a deep convolutional
architecture for NADE.
|
Rate-Distortion Bounds on Bayes Risk in Supervised Learning | We present an information-theoretic framework for bounding the number of
labeled samples needed to train a classifier in a parametric Bayesian setting.
We derive bounds on the average $L_p$ distance between the learned classifier
and the true maximum a posteriori classifier, which are well-established
surrogates for the excess classification error due to imperfect learning. We
provide lower and upper bounds on the rate-distortion function, using $L_p$
loss as the distortion measure, of a maximum a priori classifier in terms of
the differential entropy of the posterior distribution and a quantity called
the interpolation dimension, which characterizes the complexity of the
parametric distribution family. In addition to expressing the information
content of a classifier in terms of lossy compression, the rate-distortion
function also expresses the minimum number of bits a learning machine needs to
extract from training data to learn a classifier to within a specified $L_p$
tolerance. We use results from universal source coding to express the
information content in the training data in terms of the Fisher information of
the parametric family and the number of training samples available. The result
is a framework for computing lower bounds on the Bayes $L_p$ risk. This
framework complements the well-known probably approximately correct (PAC)
framework, which provides minimax risk bounds involving the Vapnik-Chervonenkis
dimension or Rademacher complexity. Whereas the PAC framework provides upper
bounds the risk for the worst-case data distribution, the proposed
rate-distortion framework lower bounds the risk averaged over the data
distribution. We evaluate the bounds for a variety of data models, including
categorical, multinomial, and Gaussian models. In each case the bounds are
provably tight orderwise, and in two cases we prove that the bounds are tight
up to multiplicative constants.
|
Predicting Performance on MOOC Assessments using Multi-Regression Models | The past few years has seen the rapid growth of data min- ing approaches for
the analysis of data obtained from Mas- sive Open Online Courses (MOOCs). The
objectives of this study are to develop approaches to predict the scores a stu-
dent may achieve on a given grade-related assessment based on information,
considered as prior performance or prior ac- tivity in the course. We develop a
personalized linear mul- tiple regression (PLMR) model to predict the grade for
a student, prior to attempting the assessment activity. The developed model is
real-time and tracks the participation of a student within a MOOC (via
click-stream server logs) and predicts the performance of a student on the next
as- sessment within the course offering. We perform a com- prehensive set of
experiments on data obtained from three openEdX MOOCs via a Stanford University
initiative. Our experimental results show the promise of the proposed ap-
proach in comparison to baseline approaches and also helps in identification of
key features that are associated with the study habits and learning behaviors
of students.
|
Active Learning for Community Detection in Stochastic Block Models | The stochastic block model (SBM) is an important generative model for random
graphs in network science and machine learning, useful for benchmarking
community detection (or clustering) algorithms. The symmetric SBM generates a
graph with $2n$ nodes which cluster into two equally sized communities. Nodes
connect with probability $p$ within a community and $q$ across different
communities. We consider the case of $p=a\ln (n)/n$ and $q=b\ln (n)/n$. In this
case, it was recently shown that recovering the community membership (or label)
of every node with high probability (w.h.p.) using only the graph is possible
if and only if the Chernoff-Hellinger (CH) divergence
$D(a,b)=(\sqrt{a}-\sqrt{b})^2 \geq 1$. In this work, we study if, and by how
much, community detection below the clustering threshold (i.e. $D(a,b)<1$) is
possible by querying the labels of a limited number of chosen nodes (i.e.,
active learning). Our main result is to show that, under certain conditions,
sampling the labels of a vanishingly small fraction of nodes (a number
sub-linear in $n$) is sufficient for exact community detection even when
$D(a,b)<1$. Furthermore, we provide an efficient learning algorithm which
recovers the community memberships of all nodes w.h.p. as long as the number of
sampled points meets the sufficient condition. We also show that recovery is
not possible if the number of observed labels is less than $n^{1-D(a,b)}$. The
validity of our results is demonstrated through numerical experiments.
|
Structured Nonconvex and Nonsmooth Optimization: Algorithms and
Iteration Complexity Analysis | Nonconvex and nonsmooth optimization problems are frequently encountered in
much of statistics, business, science and engineering, but they are not yet
widely recognized as a technology in the sense of scalability. A reason for
this relatively low degree of popularity is the lack of a well developed system
of theory and algorithms to support the applications, as is the case for its
convex counterpart. This paper aims to take one step in the direction of
disciplined nonconvex and nonsmooth optimization. In particular, we consider in
this paper some constrained nonconvex optimization models in block decision
variables, with or without coupled affine constraints. In the case of without
coupled constraints, we show a sublinear rate of convergence to an
$\epsilon$-stationary solution in the form of variational inequality for a
generalized conditional gradient method, where the convergence rate is shown to
be dependent on the H\"olderian continuity of the gradient of the smooth part
of the objective. For the model with coupled affine constraints, we introduce
corresponding $\epsilon$-stationarity conditions, and apply two proximal-type
variants of the ADMM to solve such a model, assuming the proximal ADMM updates
can be implemented for all the block variables except for the last block, for
which either a gradient step or a majorization-minimization step is
implemented. We show an iteration complexity bound of $O(1/\epsilon^2)$ to
reach an $\epsilon$-stationary solution for both algorithms. Moreover, we show
that the same iteration complexity of a proximal BCD method follows
immediately. Numerical results are provided to illustrate the efficacy of the
proposed algorithms for tensor robust PCA.
|
Randomized Kaczmarz for Rank Aggregation from Pairwise Comparisons | We revisit the problem of inferring the overall ranking among entities in the
framework of Bradley-Terry-Luce (BTL) model, based on available empirical data
on pairwise preferences. By a simple transformation, we can cast the problem as
that of solving a noisy linear system, for which a ready algorithm is available
in the form of the randomized Kaczmarz method. This scheme is provably
convergent, has excellent empirical performance, and is amenable to on-line,
distributed and asynchronous variants. Convergence, convergence rate, and error
analysis of the proposed algorithm are presented and several numerical
experiments are conducted whose results validate our theoretical findings.
|
Clustering Time Series and the Surprising Robustness of HMMs | Suppose that we are given a time series where consecutive samples are
believed to come from a probabilistic source, that the source changes from time
to time and that the total number of sources is fixed. Our objective is to
estimate the distributions of the sources. A standard approach to this problem
is to model the data as a hidden Markov model (HMM). However, since the data
often lacks the Markov or the stationarity properties of an HMM, one can ask
whether this approach is still suitable or perhaps another approach is
required. In this paper we show that a maximum likelihood HMM estimator can be
used to approximate the source distributions in a much larger class of models
than HMMs. Specifically, we propose a natural and fairly general non-stationary
model of the data, where the only restriction is that the sources do not change
too often. Our main result shows that for this model, a maximum-likelihood HMM
estimator produces the correct second moment of the data, and the results can
be extended to higher moments.
|
Random Fourier Features for Operator-Valued Kernels | Devoted to multi-task learning and structured output learning,
operator-valued kernels provide a flexible tool to build vector-valued
functions in the context of Reproducing Kernel Hilbert Spaces. To scale up
these methods, we extend the celebrated Random Fourier Feature methodology to
get an approximation of operator-valued kernels. We propose a general principle
for Operator-valued Random Fourier Feature construction relying on a
generalization of Bochner's theorem for translation-invariant operator-valued
Mercer kernels. We prove the uniform convergence of the kernel approximation
for bounded and unbounded operator random Fourier features using appropriate
Bernstein matrix concentration inequality. An experimental proof-of-concept
shows the quality of the approximation and the efficiency of the corresponding
linear models on example datasets.
|
Oracle Based Active Set Algorithm for Scalable Elastic Net Subspace
Clustering | State-of-the-art subspace clustering methods are based on expressing each
data point as a linear combination of other data points while regularizing the
matrix of coefficients with $\ell_1$, $\ell_2$ or nuclear norms. $\ell_1$
regularization is guaranteed to give a subspace-preserving affinity (i.e.,
there are no connections between points from different subspaces) under broad
theoretical conditions, but the clusters may not be connected. $\ell_2$ and
nuclear norm regularization often improve connectivity, but give a
subspace-preserving affinity only for independent subspaces. Mixed $\ell_1$,
$\ell_2$ and nuclear norm regularizations offer a balance between the
subspace-preserving and connectedness properties, but this comes at the cost of
increased computational complexity. This paper studies the geometry of the
elastic net regularizer (a mixture of the $\ell_1$ and $\ell_2$ norms) and uses
it to derive a provably correct and scalable active set method for finding the
optimal coefficients. Our geometric analysis also provides a theoretical
justification and a geometric interpretation for the balance between the
connectedness (due to $\ell_2$ regularization) and subspace-preserving (due to
$\ell_1$ regularization) properties for elastic net subspace clustering. Our
experiments show that the proposed active set method not only achieves
state-of-the-art clustering performance, but also efficiently handles
large-scale datasets.
|
Theano: A Python framework for fast computation of mathematical
expressions | Theano is a Python library that allows to define, optimize, and evaluate
mathematical expressions involving multi-dimensional arrays efficiently. Since
its introduction, it has been one of the most used CPU and GPU mathematical
compilers - especially in the machine learning community - and has shown steady
performance improvements. Theano is being actively and continuously developed
since 2008, multiple frameworks have been built on top of it and it has been
used to produce many state-of-the-art machine learning models.
The present article is structured as follows. Section I provides an overview
of the Theano software and its community. Section II presents the principal
features of Theano and how to use them, and compares them with other similar
projects. Section III focuses on recently-introduced functionalities and
improvements. Section IV compares the performance of Theano against Torch7 and
TensorFlow on several machine learning models. Section V discusses current
limitations of Theano and potential ways of improving it.
|
A Theoretical Analysis of Deep Neural Networks for Texture
Classification | We investigate the use of Deep Neural Networks for the classification of
image datasets where texture features are important for generating
class-conditional discriminative representations. To this end, we first derive
the size of the feature space for some standard textural features extracted
from the input dataset and then use the theory of Vapnik-Chervonenkis dimension
to show that hand-crafted feature extraction creates low-dimensional
representations which help in reducing the overall excess error rate. As a
corollary to this analysis, we derive for the first time upper bounds on the VC
dimension of Convolutional Neural Network as well as Dropout and Dropconnect
networks and the relation between excess error rate of Dropout and Dropconnect
networks. The concept of intrinsic dimension is used to validate the intuition
that texture-based datasets are inherently higher dimensional as compared to
handwritten digits or other object recognition datasets and hence more
difficult to be shattered by neural networks. We then derive the mean distance
from the centroid to the nearest and farthest sampling points in an
n-dimensional manifold and show that the Relative Contrast of the sample data
vanishes as dimensionality of the underlying vector space tends to infinity.
|
Nonconvex Sparse Learning via Stochastic Optimization with Progressive
Variance Reduction | We propose a stochastic variance reduced optimization algorithm for solving
sparse learning problems with cardinality constraints. Sufficient conditions
are provided, under which the proposed algorithm enjoys strong linear
convergence guarantees and optimal estimation accuracy in high dimensions. We
further extend the proposed algorithm to an asynchronous parallel variant with
a near linear speedup. Numerical experiments demonstrate the efficiency of our
algorithm in terms of both parameter estimation and computational performance.
|
LightNet: A Versatile, Standalone Matlab-based Environment for Deep
Learning | LightNet is a lightweight, versatile and purely Matlab-based deep learning
framework. The idea underlying its design is to provide an easy-to-understand,
easy-to-use and efficient computational platform for deep learning research.
The implemented framework supports major deep learning architectures such as
Multilayer Perceptron Networks (MLP), Convolutional Neural Networks (CNN) and
Recurrent Neural Networks (RNN). The framework also supports both CPU and GPU
computation, and the switch between them is straightforward. Different
applications in computer vision, natural language processing and robotics are
demonstrated as experiments.
|
Transport Analysis of Infinitely Deep Neural Network | We investigated the feature map inside deep neural networks (DNNs) by
tracking the transport map. We are interested in the role of depth (why do DNNs
perform better than shallow models?) and the interpretation of DNNs (what do
intermediate layers do?) Despite the rapid development in their application,
DNNs remain analytically unexplained because the hidden layers are nested and
the parameters are not faithful. Inspired by the integral representation of
shallow NNs, which is the continuum limit of the width, or the hidden unit
number, we developed the flow representation and transport analysis of DNNs.
The flow representation is the continuum limit of the depth or the hidden layer
number, and it is specified by an ordinary differential equation with a vector
field. We interpret an ordinary DNN as a transport map or a Euler broken line
approximation of the flow. Technically speaking, a dynamical system is a
natural model for the nested feature maps. In addition, it opens a new way to
the coordinate-free treatment of DNNs by avoiding the redundant parametrization
of DNNs. Following Wasserstein geometry, we analyze a flow in three aspects:
dynamical system, continuity equation, and Wasserstein gradient flow. A key
finding is that we specified a series of transport maps of the denoising
autoencoder (DAE). Starting from the shallow DAE, this paper develops three
topics: the transport map of the deep DAE, the equivalence between the stacked
DAE and the composition of DAEs, and the development of the double continuum
limit or the integral representation of the flow representation. As partial
answers to the research questions, we found that deeper DAEs converge faster
and the extracted features are better; in addition, a deep Gaussian DAE
transports mass to decrease the Shannon entropy of the data distribution.
|
Performance Analysis of the Gradient Comparator LMS Algorithm | The sparsity-aware zero attractor least mean square (ZA-LMS) algorithm
manifests much lower misadjustment in strongly sparse environment than its
sparsity-agnostic counterpart, the least mean square (LMS), but is shown to
perform worse than the LMS when sparsity of the impulse response decreases. The
reweighted variant of the ZA-LMS, namely RZA-LMS shows robustness against this
variation in sparsity, but at the price of increased computational complexity.
The other variants such as the l 0 -LMS and the improved proportionate
normalized LMS (IPNLMS), though perform satisfactorily, are also
computationally intensive. The gradient comparator LMS (GC-LMS) is a practical
solution of this trade-off when hardware constraint is to be considered. In
this paper, we analyse the mean and the mean square convergence performance of
the GC-LMS algorithm in detail. The analyses satisfactorily match with the
simulation results.
|
Adaptive Combination of l0 LMS Adaptive Filters for Sparse System
Identification in Fluctuating Noise Power | Recently, the l0-least mean square (l0-LMS) algorithm has been proposed to
identify sparse linear systems by employing a sparsity-promoting continuous
function as an approximation of l0 pseudonorm penalty. However, the performance
of this algorithm is sensitive to the appropriate choice of the some parameter
responsible for the zero-attracting intensity. The optimum choice for this
parameter depends on the signal-to-noise ratio (SNR) prevailing in the system.
Thus, it becomes difficult to fix a suitable value for this parameter,
particularly in a situation where SNR fluctuates over time. In this work, we
propose several adaptive combinations of differently parameterized l0-LMS to
get an overall satisfactory performance independent of the SNR, and discuss
some issues relevant to these combination structures. We also demonstrate an
efficient partial update scheme which not only reduces the number of
computations per iteration, but also achieves some interesting performance gain
compared with the full update case. Then, we propose a new recursive least
squares (RLS)-type rule to update the combining parameter more efficiently.
Finally, we extend the combination of two filters to a combination of M number
adaptive filters, which manifests further improvement for M > 2.
|
Learning theory estimates with observations from general stationary
stochastic processes | This paper investigates the supervised learning problem with observations
drawn from certain general stationary stochastic processes. Here by
\emph{general}, we mean that many stationary stochastic processes can be
included. We show that when the stochastic processes satisfy a generalized
Bernstein-type inequality, a unified treatment on analyzing the learning
schemes with various mixing processes can be conducted and a sharp oracle
inequality for generic regularized empirical risk minimization schemes can be
established. The obtained oracle inequality is then applied to derive
convergence rates for several learning schemes such as empirical risk
minimization (ERM), least squares support vector machines (LS-SVMs) using given
generic kernels, and SVMs using Gaussian kernels for both least squares and
quantile regression. It turns out that for i.i.d.~processes, our learning rates
for ERM recover the optimal rates. On the other hand, for non-i.i.d.~processes
including geometrically $\alpha$-mixing Markov processes, geometrically
$\alpha$-mixing processes with restricted decay, $\phi$-mixing processes, and
(time-reversed) geometrically $\mathcal{C}$-mixing processes, our learning
rates for SVMs with Gaussian kernels match, up to some arbitrarily small extra
term in the exponent, the optimal rates. For the remaining cases, our rates are
at least close to the optimal rates. As a by-product, the assumed generalized
Bernstein-type inequality also provides an interpretation of the so-called
"effective number of observations" for various mixing processes.
|
Web Spam Detection Using Multiple Kernels in Twin Support Vector Machine | Search engines are the most important tools for web data acquisition. Web
pages are crawled and indexed by search Engines. Users typically locate useful
web pages by querying a search engine. One of the challenges in search engines
administration is spam pages which waste search engine resources. These pages
by deception of search engine ranking algorithms try to be showed in the first
page of results. There are many approaches to web spam pages detection such as
measurement of HTML code style similarity, pages linguistic pattern analysis
and machine learning algorithm on page content features. One of the famous
algorithms has been used in machine learning approach is Support Vector Machine
(SVM) classifier. Recently basic structure of SVM has been changed by new
extensions to increase robustness and classification accuracy. In this paper we
improved accuracy of web spam detection by using two nonlinear kernels into
Twin SVM (TSVM) as an improved extension of SVM. The classifier ability to data
separation has been increased by using two separated kernels for each class of
data. Effectiveness of new proposed method has been experimented with two
publicly used spam datasets called UK-2007 and UK-2006. Results show the
effectiveness of proposed kernelized version of TSVM in web spam page
detection.
|
An efficient K-means algorithm for Massive Data | Due to the progressive growth of the amount of data available in a wide
variety of scientific fields, it has become more difficult to ma- nipulate and
analyze such information. Even though datasets have grown in size, the K-means
algorithm remains as one of the most popular clustering methods, in spite of
its dependency on the initial settings and high computational cost, especially
in terms of distance computations. In this work, we propose an efficient
approximation to the K-means problem intended for massive data. Our approach
recursively partitions the entire dataset into a small number of sub- sets,
each of which is characterized by its representative (center of mass) and
weight (cardinality), afterwards a weighted version of the K-means algorithm is
applied over such local representation, which can drastically reduce the number
of distances computed. In addition to some theoretical properties, experimental
results indicate that our method outperforms well-known approaches, such as the
K-means++ and the minibatch K-means, in terms of the relation between number of
distance computations and the quality of the approximation.
|
MUST-CNN: A Multilayer Shift-and-Stitch Deep Convolutional Architecture
for Sequence-based Protein Structure Prediction | Predicting protein properties such as solvent accessibility and secondary
structure from its primary amino acid sequence is an important task in
bioinformatics. Recently, a few deep learning models have surpassed the
traditional window based multilayer perceptron. Taking inspiration from the
image classification domain we propose a deep convolutional neural network
architecture, MUST-CNN, to predict protein properties. This architecture uses a
novel multilayer shift-and-stitch (MUST) technique to generate fully dense
per-position predictions on protein sequences. Our model is significantly
simpler than the state-of-the-art, yet achieves better results. By combining
MUST and the efficient convolution operation, we can consider far more
parameters while retaining very fast prediction speeds. We beat the
state-of-the-art performance on two large protein property prediction datasets.
|
Semi-Supervised Representation Learning based on Probabilistic Labeling | In this paper, we present a new algorithm for semi-supervised representation
learning. In this algorithm, we first find a vector representation for the
labels of the data points based on their local positions in the space. Then, we
map the data to lower-dimensional space using a linear transformation such that
the dependency between the transformed data and the assigned labels is
maximized. In fact, we try to find a mapping that is as discriminative as
possible. The approach will use Hilber-Schmidt Independence Criterion (HSIC) as
the dependence measure. We also present a kernelized version of the algorithm,
which allows non-linear transformations and provides more flexibility in
finding the appropriate mapping. Use of unlabeled data for learning new
representation is not always beneficial and there is no algorithm that can
deterministically guarantee the improvement of the performance by exploiting
unlabeled data. Therefore, we also propose a bound on the performance of the
algorithm, which can be used to determine the effectiveness of using the
unlabeled data in the algorithm. We demonstrate the ability of the algorithm in
finding the transformation using both toy examples and real-world datasets.
|
Active Uncertainty Calibration in Bayesian ODE Solvers | There is resurging interest, in statistics and machine learning, in solvers
for ordinary differential equations (ODEs) that return probability measures
instead of point estimates. Recently, Conrad et al. introduced a sampling-based
class of methods that are 'well-calibrated' in a specific sense. But the
computational cost of these methods is significantly above that of classic
methods. On the other hand, Schober et al. pointed out a precise connection
between classic Runge-Kutta ODE solvers and Gaussian filters, which gives only
a rough probabilistic calibration, but at negligible cost overhead. By
formulating the solution of ODEs as approximate inference in linear Gaussian
SDEs, we investigate a range of probabilistic ODE solvers, that bridge the
trade-off between computational cost and probabilistic calibration, and
identify the inaccurate gradient measurement as the crucial source of
uncertainty. We propose the novel filtering-based method Bayesian Quadrature
filtering (BQF) which uses Bayesian quadrature to actively learn the
imprecision in the gradient measurement by collecting multiple gradient
evaluations.
|
Unbiased split variable selection for random survival forests using
maximally selected rank statistics | The most popular approach for analyzing survival data is the Cox regression
model. The Cox model may, however, be misspecified, and its proportionality
assumption may not always be fulfilled. An alternative approach for survival
prediction is random forests for survival outcomes. The standard split
criterion for random survival forests is the log-rank test statistics, which
favors splitting variables with many possible split points. Conditional
inference forests avoid this split variable selection bias. However, linear
rank statistics are utilized by default in conditional inference forests to
select the optimal splitting variable, which cannot detect non-linear effects
in the independent variables. An alternative is to use maximally selected rank
statistics for the split point selection. As in conditional inference forests,
splitting variables are compared on the p-value scale. However, instead of the
conditional Monte-Carlo approach used in conditional inference forests, p-value
approximations are employed. We describe several p-value approximations and the
implementation of the proposed random forest approach. A simulation study
demonstrates that unbiased split variable selection is possible. However, there
is a trade-off between unbiased split variable selection and runtime. In
benchmark studies of prediction performance on simulated and real datasets the
new method performs better than random survival forests if informative
dichotomous variables are combined with uninformative variables with more
categories and better than conditional inference forests if non-linear
covariate effects are included. In a runtime comparison the method proves to be
computationally faster than both alternatives, if a simple p-value
approximation is used.
|
A constrained L1 minimization approach for estimating multiple Sparse
Gaussian or Nonparanormal Graphical Models | Identifying context-specific entity networks from aggregated data is an
important task, arising often in bioinformatics and neuroimaging.
Computationally, this task can be formulated as jointly estimating multiple
different, but related, sparse Undirected Graphical Models (UGM) from
aggregated samples across several contexts. Previous joint-UGM studies have
mostly focused on sparse Gaussian Graphical Models (sGGMs) and can't identify
context-specific edge patterns directly. We, therefore, propose a novel
approach, SIMULE (detecting Shared and Individual parts of MULtiple graphs
Explicitly) to learn multi-UGM via a constrained L1 minimization. SIMULE
automatically infers both specific edge patterns that are unique to each
context and shared interactions preserved among all the contexts. Through the
L1 constrained formulation, this problem is cast as multiple independent
subtasks of linear programming that can be solved efficiently in parallel. In
addition to Gaussian data, SIMULE can also handle multivariate Nonparanormal
data that greatly relaxes the normality assumption that many real-world
applications do not follow. We provide a novel theoretical proof showing that
SIMULE achieves a consistent result at the rate O(log(Kp)/n_{tot}). On multiple
synthetic datasets and two biomedical datasets, SIMULE shows significant
improvement over state-of-the-art multi-sGGM and single-UGM baselines.
|
Tweet2Vec: Character-Based Distributed Representations for Social Media | Text from social media provides a set of challenges that can cause
traditional NLP approaches to fail. Informal language, spelling errors,
abbreviations, and special characters are all commonplace in these posts,
leading to a prohibitively large vocabulary size for word-level approaches. We
propose a character composition model, tweet2vec, which finds vector-space
representations of whole tweets by learning complex, non-local dependencies in
character sequences. The proposed model outperforms a word-level baseline at
predicting user-annotated hashtags associated with the posts, doing
significantly better when the input contains many out-of-vocabulary words or
unusual character sequences. Our tweet2vec encoder is publicly available.
|
On the Iteration Complexity of Oblivious First-Order Optimization
Algorithms | We consider a broad class of first-order optimization algorithms which are
\emph{oblivious}, in the sense that their step sizes are scheduled regardless
of the function under consideration, except for limited side-information such
as smoothness or strong convexity parameters. With the knowledge of these two
parameters, we show that any such algorithm attains an iteration complexity
lower bound of $\Omega(\sqrt{L/\epsilon})$ for $L$-smooth convex functions, and
$\tilde{\Omega}(\sqrt{L/\mu}\ln(1/\epsilon))$ for $L$-smooth $\mu$-strongly
convex functions. These lower bounds are stronger than those in the traditional
oracle model, as they hold independently of the dimension. To attain these, we
abandon the oracle model in favor of a structure-based approach which builds
upon a framework recently proposed in (Arjevani et al., 2015). We further show
that without knowing the strong convexity parameter, it is impossible to attain
an iteration complexity better than
$\tilde{\Omega}\left((L/\mu)\ln(1/\epsilon)\right)$. This result is then used
to formalize an observation regarding $L$-smooth convex functions, namely, that
the iteration complexity of algorithms employing time-invariant step sizes must
be at least $\Omega(L/\epsilon)$.
|
EEF: Exponentially Embedded Families with Class-Specific Features for
Classification | In this letter, we present a novel exponentially embedded families (EEF)
based classification method, in which the probability density function (PDF) on
raw data is estimated from the PDF on features. With the PDF construction, we
show that class-specific features can be used in the proposed classification
method, instead of a common feature subset for all classes as used in
conventional approaches. We apply the proposed EEF classifier for text
categorization as a case study and derive an optimal Bayesian classification
rule with class-specific feature selection based on the Information Gain (IG)
score. The promising performance on real-life data sets demonstrates the
effectiveness of the proposed approach and indicates its wide potential
applications.
|
Learning Representations for Counterfactual Inference | Observational studies are rising in importance due to the widespread
accumulation of data in fields such as healthcare, education, employment and
ecology. We consider the task of answering counterfactual questions such as,
"Would this patient have lower blood sugar had she received a different
medication?". We propose a new algorithmic framework for counterfactual
inference which brings together ideas from domain adaptation and representation
learning. In addition to a theoretical justification, we perform an empirical
comparison with previous approaches to causal inference from observational
data. Our deep learning algorithm significantly outperforms the previous
state-of-the-art.
|
Exponential Machines | Modeling interactions between features improves the performance of machine
learning solutions in many domains (e.g. recommender systems or sentiment
analysis). In this paper, we introduce Exponential Machines (ExM), a predictor
that models all interactions of every order. The key idea is to represent an
exponentially large tensor of parameters in a factorized format called Tensor
Train (TT). The Tensor Train format regularizes the model and lets you control
the number of underlying parameters. To train the model, we develop a
stochastic Riemannian optimization procedure, which allows us to fit tensors
with 2^160 entries. We show that the model achieves state-of-the-art
performance on synthetic data with high-order interactions and that it works on
par with high-order factorization machines on a recommender system dataset
MovieLens 100K.
|
Detecting Relative Anomaly | System states that are anomalous from the perspective of a domain expert
occur frequently in some anomaly detection problems. The performance of
commonly used unsupervised anomaly detection methods may suffer in that
setting, because they use frequency as a proxy for anomaly. We propose a novel
concept for anomaly detection, called relative anomaly detection. It is
tailored to be robust towards anomalies that occur frequently, by taking into
account their location relative to the most typical observations. The
approaches we develop are computationally feasible even for large data sets,
and they allow real-time detection. We illustrate using data sets of potential
scraping attempts and Wi-Fi channel utilization, both from Google, Inc.
|
Noisy Parallel Approximate Decoding for Conditional Recurrent Language
Model | Recent advances in conditional recurrent language modelling have mainly
focused on network architectures (e.g., attention mechanism), learning
algorithms (e.g., scheduled sampling and sequence-level training) and novel
applications (e.g., image/video description generation, speech recognition,
etc.) On the other hand, we notice that decoding algorithms/strategies have not
been investigated as much, and it has become standard to use greedy or beam
search. In this paper, we propose a novel decoding strategy motivated by an
earlier observation that nonlinear hidden layers of a deep neural network
stretch the data manifold. The proposed strategy is embarrassingly
parallelizable without any communication overhead, while improving an existing
decoding algorithm. We extensively evaluate it with attention-based neural
machine translation on the task of En->Cz translation.
|
Asymptotic sequential Rademacher complexity of a finite function class | For a finite function class we describe the large sample limit of the
sequential Rademacher complexity in terms of the viscosity solution of a
$G$-heat equation. In the language of Peng's sublinear expectation theory, the
same quantity equals to the expected value of the largest order statistics of a
multidimensional $G$-normal random variable. We illustrate this result by
deriving upper and lower bounds for the asymptotic sequential Rademacher
complexity.
|
Context-dependent feature analysis with random forests | In many cases, feature selection is often more complicated than identifying a
single subset of input variables that would together explain the output. There
may be interactions that depend on contextual information, i.e., variables that
reveal to be relevant only in some specific circumstances. In this setting, the
contribution of this paper is to extend the random forest variable importances
framework in order (i) to identify variables whose relevance is
context-dependent and (ii) to characterize as precisely as possible the effect
of contextual information on these variables. The usage and the relevance of
our framework for highlighting context-dependent variables is illustrated on
both artificial and real datasets.
|
Competitive analysis of the top-K ranking problem | Motivated by applications in recommender systems, web search, social choice
and crowdsourcing, we consider the problem of identifying the set of top $K$
items from noisy pairwise comparisons. In our setting, we are non-actively
given $r$ pairwise comparisons between each pair of $n$ items, where each
comparison has noise constrained by a very general noise model called the
strong stochastic transitivity (SST) model. We analyze the competitive ratio of
algorithms for the top-$K$ problem. In particular, we present a linear time
algorithm for the top-$K$ problem which has a competitive ratio of
$\tilde{O}(\sqrt{n})$; i.e. to solve any instance of top-$K$, our algorithm
needs at most $\tilde{O}(\sqrt{n})$ times as many samples needed as the best
possible algorithm for that instance (in contrast, all previous known
algorithms for the top-$K$ problem have competitive ratios of
$\tilde{\Omega}(n)$ or worse). We further show that this is tight: any
algorithm for the top-$K$ problem has competitive ratio at least
$\tilde{\Omega}(\sqrt{n})$.
|
Transfer Hashing with Privileged Information | Most existing learning to hash methods assume that there are sufficient data,
either labeled or unlabeled, on the domain of interest (i.e., the target
domain) for training. However, this assumption cannot be satisfied in some
real-world applications. To address this data sparsity issue in hashing,
inspired by transfer learning, we propose a new framework named Transfer
Hashing with Privileged Information (THPI). Specifically, we extend the
standard learning to hash method, Iterative Quantization (ITQ), in a transfer
learning manner, namely ITQ+. In ITQ+, a new slack function is learned from
auxiliary data to approximate the quantization error in ITQ. We developed an
alternating optimization approach to solve the resultant optimization problem
for ITQ+. We further extend ITQ+ to LapITQ+ by utilizing the geometry structure
among the auxiliary data for learning more precise binary codes in the target
domain. Extensive experiments on several benchmark datasets verify the
effectiveness of our proposed approaches through comparisons with several
state-of-the-art baselines.
|
Causal Discovery for Manufacturing Domains | Yield and quality improvement is of paramount importance to any manufacturing
company. One of the ways of improving yield is through discovery of the root
causal factors affecting yield. We propose the use of data-driven interpretable
causal models to identify key factors affecting yield. We focus on factors that
are measured in different stages of production and testing in the manufacturing
cycle of a product. We apply causal structure learning techniques on real data
collected from this line. Specifically, the goal of this work is to learn
interpretable causal models from observational data produced by manufacturing
lines.
Emphasis has been given to the interpretability of the models to make them
actionable in the field of manufacturing. We highlight the challenges presented
by assembly line data and propose ways to alleviate them.We also identify
unique characteristics of data originating from assembly lines and how to
leverage them in order to improve causal discovery. Standard evaluation
techniques for causal structure learning shows that the learned causal models
seem to closely represent the underlying latent causal relationship between
different factors in the production process. These results were also validated
by manufacturing domain experts who found them promising. This work
demonstrates how data mining and knowledge discovery can be used for root cause
analysis in the domain of manufacturing and connected industry.
|
A Reinforcement Learning System to Encourage Physical Activity in
Diabetes Patients | Regular physical activity is known to be beneficial to people suffering from
diabetes type 2. Nevertheless, most such people are sedentary. Smartphones
create new possibilities for helping people to adhere to their physical
activity goals, through continuous monitoring and communication, coupled with
personalized feedback.
We provided 27 sedentary diabetes type 2 patients with a smartphone-based
pedometer and a personal plan for physical activity. Patients were sent SMS
messages to encourage physical activity between once a day and once per week.
Messages were personalized through a Reinforcement Learning (RL) algorithm
which optimized messages to improve each participant's compliance with the
activity regimen. The RL algorithm was compared to a static policy for sending
messages and to weekly reminders.
Our results show that participants who received messages generated by the RL
algorithm increased the amount of activity and pace of walking, while the
control group patients did not. Patients assigned to the RL algorithm group
experienced a superior reduction in blood glucose levels (HbA1c) compared to
control policies, and longer participation caused greater reductions in blood
glucose levels. The learning algorithm improved gradually in predicting which
messages would lead participants to exercise.
Our results suggest that a mobile phone application coupled with a learning
algorithm can improve adherence to exercise in diabetic patients. As a learning
algorithm is automated, and delivers personalized messages, it could be used in
large populations of diabetic patients to improve health and glycemic control.
Our results can be expanded to other areas where computer-led health coaching
of humans may have a positive impact.
|
Barzilai-Borwein Step Size for Stochastic Gradient Descent | One of the major issues in stochastic gradient descent (SGD) methods is how
to choose an appropriate step size while running the algorithm. Since the
traditional line search technique does not apply for stochastic optimization
algorithms, the common practice in SGD is either to use a diminishing step
size, or to tune a fixed step size by hand, which can be time consuming in
practice. In this paper, we propose to use the Barzilai-Borwein (BB) method to
automatically compute step sizes for SGD and its variant: stochastic variance
reduced gradient (SVRG) method, which leads to two algorithms: SGD-BB and
SVRG-BB. We prove that SVRG-BB converges linearly for strongly convex objective
functions. As a by-product, we prove the linear convergence result of SVRG with
Option I proposed in [10], whose convergence result is missing in the
literature. Numerical experiments on standard data sets show that the
performance of SGD-BB and SVRG-BB is comparable to and sometimes even better
than SGD and SVRG with best-tuned step sizes, and is superior to some advanced
SGD variants.
|
Online Optimization Methods for the Quantification Problem | The estimation of class prevalence, i.e., the fraction of a population that
belongs to a certain class, is a very useful tool in data analytics and
learning, and finds applications in many domains such as sentiment analysis,
epidemiology, etc. For example, in sentiment analysis, the objective is often
not to estimate whether a specific text conveys a positive or a negative
sentiment, but rather estimate the overall distribution of positive and
negative sentiments during an event window. A popular way of performing the
above task, often dubbed quantification, is to use supervised learning to train
a prevalence estimator from labeled data.
Contemporary literature cites several performance measures used to measure
the success of such prevalence estimators. In this paper we propose the first
online stochastic algorithms for directly optimizing these
quantification-specific performance measures. We also provide algorithms that
optimize hybrid performance measures that seek to balance quantification and
classification performance. Our algorithms present a significant advancement in
the theory of multivariate optimization and we show, by a rigorous theoretical
analysis, that they exhibit optimal convergence. We also report extensive
experiments on benchmark and real data sets which demonstrate that our methods
significantly outperform existing optimization techniques used for these
performance measures.
|
Support Vector Algorithms for Optimizing the Partial Area Under the ROC
Curve | The area under the ROC curve (AUC) is a widely used performance measure in
machine learning. Increasingly, however, in several applications, ranging from
ranking to biometric screening to medicine, performance is measured not in
terms of the full area under the ROC curve, but in terms of the \emph{partial}
area under the ROC curve between two false positive rates. In this paper, we
develop support vector algorithms for directly optimizing the partial AUC
between any two false positive rates. Our methods are based on minimizing a
suitable proxy or surrogate objective for the partial AUC error. In the case of
the full AUC, one can readily construct and optimize convex surrogates by
expressing the performance measure as a summation of pairwise terms. The
partial AUC, on the other hand, does not admit such a simple decomposable
structure, making it more challenging to design and optimize (tight) convex
surrogates for this measure.
Our approach builds on the structural SVM framework of Joachims (2005) to
design convex surrogates for partial AUC, and solves the resulting optimization
problem using a cutting plane solver. Unlike the full AUC, where the
combinatorial optimization needed in each iteration of the cutting plane solver
can be decomposed and solved efficiently, the corresponding problem for the
partial AUC is harder to decompose. One of our main contributions is a
polynomial time algorithm for solving the combinatorial optimization problem
associated with partial AUC. We also develop an approach for optimizing a
tighter non-convex hinge loss based surrogate for the partial AUC using
difference-of-convex programming. Our experiments on a variety of real-world
and benchmark tasks confirm the efficacy of the proposed methods.
|
Monotone Retargeting for Unsupervised Rank Aggregation with Object
Features | Learning the true ordering between objects by aggregating a set of expert
opinion rank order lists is an important and ubiquitous problem in many
applications ranging from social choice theory to natural language processing
and search aggregation. We study the problem of unsupervised rank aggregation
where no ground truth ordering information in available, neither about the true
preference ordering between any set of objects nor about the quality of
individual rank lists. Aggregating the often inconsistent and poor quality rank
lists in such an unsupervised manner is a highly challenging problem, and
standard consensus-based methods are often ill-defined, and difficult to solve.
In this manuscript we propose a novel framework to bypass these issues by using
object attributes to augment the standard rank aggregation framework. We design
algorithms that learn joint models on both rank lists and object features to
obtain an aggregated rank ordering that is more accurate and robust, and also
helps weed out rank lists of dubious validity. We validate our techniques on
synthetic datasets where our algorithm is able to estimate the true rank
ordering even when the rank lists are corrupted. Experiments on three real
datasets, MQ2008, MQ2008 and OHSUMED, show that using object features can
result in significant improvement in performance over existing rank aggregation
methods that do not use object information. Furthermore, when at least some of
the rank lists are of high quality, our methods are able to effectively exploit
their high expertise to output an aggregated rank ordering of great accuracy.
|
Generalized Linear Models for Aggregated Data | Databases in domains such as healthcare are routinely released to the public
in aggregated form. Unfortunately, naive modeling with aggregated data may
significantly diminish the accuracy of inferences at the individual level. This
paper addresses the scenario where features are provided at the individual
level, but the target variables are only available as histogram aggregates or
order statistics. We consider a limiting case of generalized linear modeling
when the target variables are only known up to permutation, and explore how
this relates to permutation testing; a standard technique for assessing
statistical dependency. Based on this relationship, we propose a simple
algorithm to estimate the model parameters and individual level inferences via
alternating imputation and standard generalized linear model fitting. Our
results suggest the effectiveness of the proposed approach when, in the
original data, permutation testing accurately ascertains the veracity of the
linear relationship. The framework is extended to general histogram data with
larger bins - with order statistics such as the median as a limiting case. Our
experimental results on simulated data and aggregated healthcare data suggest a
diminishing returns property with respect to the granularity of the histogram -
when a linear relationship holds in the original data, the targets can be
predicted accurately given relatively coarse histograms.
|
DeepLearningKit - an GPU Optimized Deep Learning Framework for Apple's
iOS, OS X and tvOS developed in Metal and Swift | In this paper we present DeepLearningKit - an open source framework that
supports using pretrained deep learning models (convolutional neural networks)
for iOS, OS X and tvOS. DeepLearningKit is developed in Metal in order to
utilize the GPU efficiently and Swift for integration with applications, e.g.
iOS-based mobile apps on iPhone/iPad, tvOS-based apps for the big screen, or OS
X desktop applications. The goal is to support using deep learning models
trained with popular frameworks such as Caffe, Torch, TensorFlow, Theano,
Pylearn, Deeplearning4J and Mocha. Given the massive GPU resources and time
required to train Deep Learning models we suggest an App Store like model to
distribute and download pretrained and reusable Deep Learning models.
|
Learning to Rank Personalized Search Results in Professional Networks | LinkedIn search is deeply personalized - for the same queries, different
searchers expect completely different results. This paper presents our approach
to achieving this by mining various data sources available in LinkedIn to infer
searchers' intents (such as hiring, job seeking, etc.), as well as extending
the concept of homophily to capture the searcher-result similarities on many
aspects. Then, learning-to-rank (LTR) is applied to combine these signals with
standard search features.
|
Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online
Learning with True and Noisy Gradient | This work focuses on dynamic regret of online convex optimization that
compares the performance of online learning to a clairvoyant who knows the
sequence of loss functions in advance and hence selects the minimizer of the
loss function at each step. By assuming that the clairvoyant moves slowly
(i.e., the minimizers change slowly), we present several improved
variation-based upper bounds of the dynamic regret under the true and noisy
gradient feedback, which are {\it optimal} in light of the presented lower
bounds. The key to our analysis is to explore a regularity metric that measures
the temporal changes in the clairvoyant's minimizers, to which we refer as {\it
path variation}. Firstly, we present a general lower bound in terms of the path
variation, and then show that under full information or gradient feedback we
are able to achieve an optimal dynamic regret. Secondly, we present a lower
bound with noisy gradient feedback and then show that we can achieve optimal
dynamic regrets under a stochastic gradient feedback and two-point bandit
feedback. Moreover, for a sequence of smooth loss functions that admit a small
variation in the gradients, our dynamic regret under the two-point bandit
feedback matches what is achieved with full information.
|
Alternating optimization method based on nonnegative matrix
factorizations for deep neural networks | The backpropagation algorithm for calculating gradients has been widely used
in computation of weights for deep neural networks (DNNs). This method requires
derivatives of objective functions and has some difficulties finding
appropriate parameters such as learning rate. In this paper, we propose a novel
approach for computing weight matrices of fully-connected DNNs by using two
types of semi-nonnegative matrix factorizations (semi-NMFs). In this method,
optimization processes are performed by calculating weight matrices
alternately, and backpropagation (BP) is not used. We also present a method to
calculate stacked autoencoder using a NMF. The output results of the
autoencoder are used as pre-training data for DNNs. The experimental results
show that our method using three types of NMFs attains similar error rates to
the conventional DNNs with BP.
|
Fast and Accurate Performance Analysis of LTE Radio Access Networks | An increasing amount of analytics is performed on data that is procured in a
real-time fashion to make real-time decisions. Such tasks include simple
reporting on streams to sophisticated model building. However, the practicality
of such analyses are impeded in several domains because they are faced with a
fundamental trade-off between data collection latency and analysis accuracy.
In this paper, we study this trade-off in the context of a specific domain,
Cellular Radio Access Networks (RAN). Our choice of this domain is influenced
by its commonalities with several other domains that produce real-time data,
our access to a large live dataset, and their real-time nature and
dimensionality which makes it a natural fit for a popular analysis technique,
machine learning (ML). We find that the latency accuracy trade-off can be
resolved using two broad, general techniques: intelligent data grouping and
task formulations that leverage domain characteristics. Based on this, we
present CellScope, a system that addresses this challenge by applying a domain
specific formulation and application of Multi-task Learning (MTL) to RAN
performance analysis. It achieves this goal using three techniques: feature
engineering to transform raw data into effective features, a PCA inspired
similarity metric to group data from geographically nearby base stations
sharing performance commonalities, and a hybrid online-offline model for
efficient model updates. Our evaluation of CellScope shows that its accuracy
improvements over direct application of ML range from 2.5x to 4.4x while
reducing the model update overhead by up to 4.8x. We have also used CellScope
to analyze a live LTE consisting of over 2 million subscribers for a period of
over 10 months, where it uncovered several problems and insights, some of them
previously unknown.
|
Joint Learning of Sentence Embeddings for Relevance and Entailment | We consider the problem of Recognizing Textual Entailment within an
Information Retrieval context, where we must simultaneously determine the
relevancy as well as degree of entailment for individual pieces of evidence to
determine a yes/no answer to a binary natural language question.
We compare several variants of neural networks for sentence embeddings in a
setting of decision-making based on evidence of varying relevance. We propose a
basic model to integrate evidence for entailment, show that joint training of
the sentence embeddings to model relevance and entailment is feasible even with
no explicit per-evidence supervision, and show the importance of evaluating
strong baselines. We also demonstrate the benefit of carrying over text
comprehension model trained on an unrelated task for our small datasets.
Our research is motivated primarily by a new open dataset we introduce,
consisting of binary questions and news-based evidence snippets. We also apply
the proposed relevance-entailment model on a similar task of ranking
multiple-choice test answers, evaluating it on a preliminary dataset of school
test questions as well as the standard MCTest dataset, where we improve the
neural model state-of-art.
|
Solve-Select-Scale: A Three Step Process For Sparse Signal Estimation | In the theory of compressed sensing (CS), the sparsity $\|x\|_0$ of the
unknown signal $\mathbf{x} \in \mathcal{R}^n$ is of prime importance and the
focus of reconstruction algorithms has mainly been either $\|x\|_0$ or its
convex relaxation (via $\|x\|_1$). However, it is typically unknown in practice
and has remained a challenge when nothing about the size of the support is
known. As pointed recently, $\|x\|_0$ might not be the best metric to minimize
directly, both due to its inherent complexity as well as its noise performance.
Recently a novel stable measure of sparsity $s(\mathbf{x}) :=
\|\mathbf{x}\|_1^2/\|\mathbf{x}\|_2^2$ has been investigated by Lopes
\cite{Lopes2012}, which is a sharp lower bound on $\|\mathbf{x}\|_0$. The
estimation procedure for this measure uses only a small number of linear
measurements, does not rely on any sparsity assumptions, and requires very
little computation. The usage of the quantity $s(\mathbf{x})$ in sparse signal
estimation problems has not received much importance yet. We develop the idea
of incorporating $s(\mathbf{x})$ into the signal estimation framework. We also
provide a three step algorithm to solve problems of the form $\mathbf{Ax=b}$
with no additional assumptions on the original signal $\mathbf{x}$.
|
A Critical Examination of RESCAL for Completion of Knowledge Bases with
Transitive Relations | Link prediction in large knowledge graphs has received a lot of attention
recently because of its importance for inferring missing relations and for
completing and improving noisily extracted knowledge graphs. Over the years a
number of machine learning researchers have presented various models for
predicting the presence of missing relations in a knowledge base. Although all
the previous methods are presented with empirical results that show high
performance on select datasets, there is almost no previous work on
understanding the connection between properties of a knowledge base and the
performance of a model. In this paper we analyze the RESCAL method and prove
that it can not encode asymmetric transitive relations in knowledge bases.
|
CNN based texture synthesize with Semantic segment | Deep learning algorithm display powerful ability in Computer Vision area, in
recent year, the CNN has been applied to solve problems in the subarea of
Image-generating, which has been widely applied in areas such as photo editing,
image design, computer animation, real-time rendering for large scale of scenes
and for visual effects in movies. However in the texture synthesize procedure.
The state-of-art CNN can not capture the spatial location of texture in image,
lead to significant distortion after texture synthesize, we propose a new way
to generating-image by adding the semantic segment step with deep learning
algorithm as Pre-Processing and analyze the outcome.
|
Geometry Aware Mappings for High Dimensional Sparse Factors | While matrix factorisation models are ubiquitous in large scale
recommendation and search, real time application of such models requires inner
product computations over an intractably large set of item factors. In this
manuscript we present a novel framework that uses the inverted index
representation to exploit structural properties of sparse vectors to
significantly reduce the run time computational cost of factorisation models.
We develop techniques that use geometry aware permutation maps on a tessellated
unit sphere to obtain high dimensional sparse embeddings for latent factors
with sparsity patterns related to angular closeness of the original latent
factors. We also design several efficient and deterministic realisations within
this framework and demonstrate with experiments that our techniques lead to
faster run time operation with minimal loss of accuracy.
|
Off-policy evaluation for slate recommendation | This paper studies the evaluation of policies that recommend an ordered set
of items (e.g., a ranking) based on some context---a common scenario in web
search, ads, and recommendation. We build on techniques from combinatorial
bandits to introduce a new practical estimator that uses logged data to
estimate a policy's performance. A thorough empirical evaluation on real-world
data reveals that our estimator is accurate in a variety of settings, including
as a subroutine in a learning-to-rank task, where it achieves competitive
performance. We derive conditions under which our estimator is unbiased---these
conditions are weaker than prior heuristics for slate evaluation---and
experimentally demonstrate a smaller bias than parametric approaches, even when
these conditions are violated. Finally, our theory and experiments also show
exponential savings in the amount of required data compared with general
unbiased estimators.
|
Reducing the Model Order of Deep Neural Networks Using Information
Theory | Deep neural networks are typically represented by a much larger number of
parameters than shallow models, making them prohibitive for small footprint
devices. Recent research shows that there is considerable redundancy in the
parameter space of deep neural networks. In this paper, we propose a method to
compress deep neural networks by using the Fisher Information metric, which we
estimate through a stochastic optimization method that keeps track of
second-order information in the network. We first remove unimportant parameters
and then use non-uniform fixed point quantization to assign more bits to
parameters with higher Fisher Information estimates. We evaluate our method on
a classification task with a convolutional neural network trained on the MNIST
data set. Experimental results show that our method outperforms existing
methods for both network pruning and quantization.
|
Gearbox Fault Detection through PSO Exact Wavelet Analysis and SVM
Classifier | Time-frequency methods for vibration-based gearbox faults detection have been
considered the most efficient method. Among these methods, continuous wavelet
transform (CWT) as one of the best time-frequency method has been used for both
stationary and transitory signals. Some deficiencies of CWT are problem of
overlapping and distortion ofsignals. In this condition, a large amount of
redundant information exists so that it may cause false alarm or
misinterpretation of the operator. In this paper a modified method called Exact
Wavelet Analysis is used to minimize the effects of overlapping and distortion
in case of gearbox faults. To implement exact wavelet analysis, Particle Swarm
Optimization (PSO) algorithm has been used for this purpose. This method have
been implemented for the acceleration signals from 2D acceleration sensor
acquired by Advantech PCI-1710 card from a gearbox test setup in Amirkabir
University of Technology. Gearbox has been considered in both healthy and
chipped tooth gears conditions. Kernelized Support Vector Machine (SVM) with
radial basis functions has used the extracted features from exact wavelet
analysis for classification. The efficiency of this classifier is then
evaluated with the other signals acquired from the setup test. The results show
that in comparison of CWT, PSO Exact Wavelet Transform has better ability in
feature extraction in price of more computational effort. In addition, PSO
exact wavelet has better speed comparing to Genetic Algorithm (GA) exact
wavelet in condition of equal population because of factoring mutation and
crossover in PSO algorithm. SVM classifier with the extracted features in
gearbox shows very good results and its ability has been proved.
|
A Constant-Factor Bi-Criteria Approximation Guarantee for $k$-means++ | This paper studies the $k$-means++ algorithm for clustering as well as the
class of $D^\ell$ sampling algorithms to which $k$-means++ belongs. It is shown
that for any constant factor $\beta > 1$, selecting $\beta k$ cluster centers
by $D^\ell$ sampling yields a constant-factor approximation to the optimal
clustering with $k$ centers, in expectation and without conditions on the
dataset. This result extends the previously known $O(\log k)$ guarantee for the
case $\beta = 1$ to the constant-factor bi-criteria regime. It also improves
upon an existing constant-factor bi-criteria result that holds only with
constant probability.
|
Locally Weighted Ensemble Clustering | Due to its ability to combine multiple base clusterings into a probably
better and more robust clustering, the ensemble clustering technique has been
attracting increasing attention in recent years. Despite the significant
success, one limitation to most of the existing ensemble clustering methods is
that they generally treat all base clusterings equally regardless of their
reliability, which makes them vulnerable to low-quality base clusterings.
Although some efforts have been made to (globally) evaluate and weight the base
clusterings, yet these methods tend to view each base clustering as an
individual and neglect the local diversity of clusters inside the same base
clustering. It remains an open problem how to evaluate the reliability of
clusters and exploit the local diversity in the ensemble to enhance the
consensus performance, especially in the case when there is no access to data
features or specific assumptions on data distribution. To address this, in this
paper, we propose a novel ensemble clustering approach based on ensemble-driven
cluster uncertainty estimation and local weighting strategy. In particular, the
uncertainty of each cluster is estimated by considering the cluster labels in
the entire ensemble via an entropic criterion. A novel ensemble-driven cluster
validity measure is introduced, and a locally weighted co-association matrix is
presented to serve as a summary for the ensemble of diverse clusters. With the
local diversity in ensembles exploited, two novel consensus functions are
further proposed. Extensive experiments on a variety of real-world datasets
demonstrate the superiority of the proposed approach over the state-of-the-art.
|
Incremental Robot Learning of New Objects with Fixed Update Time | We consider object recognition in the context of lifelong learning, where a
robotic agent learns to discriminate between a growing number of object classes
as it accumulates experience about the environment. We propose an incremental
variant of the Regularized Least Squares for Classification (RLSC) algorithm,
and exploit its structure to seamlessly add new classes to the learned model.
The presented algorithm addresses the problem of having an unbalanced
proportion of training examples per class, which occurs when new objects are
presented to the system for the first time.
We evaluate our algorithm on both a machine learning benchmark dataset and
two challenging object recognition tasks in a robotic setting. Empirical
evidence shows that our approach achieves comparable or higher classification
performance than its batch counterpart when classes are unbalanced, while being
significantly faster.
|
Word2Vec is a special case of Kernel Correspondence Analysis and Kernels
for Natural Language Processing | We show that correspondence analysis (CA) is equivalent to defining a Gini
index with appropriately scaled one-hot encoding. Using this relation, we
introduce a nonlinear kernel extension to CA. This extended CA gives a known
analysis for natural language via specialized kernels that use an appropriate
contingency table. We propose a semi-supervised CA, which is a special case of
the kernel extension to CA. Because CA requires excessive memory if applied to
numerous categories, CA has not been used for natural language processing. We
address this problem by introducing delayed evaluation to randomized singular
value decomposition. The memory-efficient CA is then applied to a word-vector
representation task. We propose a tail-cut kernel, which is an extension to the
skip-gram within the kernel extension to CA. Our tail-cut kernel outperforms
existing word-vector representation methods.
|
Automatic Classification of Irregularly Sampled Time Series with Unequal
Lengths: A Case Study on Estimated Glomerular Filtration Rate | A patient's estimated glomerular filtration rate (eGFR) can provide important
information about disease progression and kidney function. Traditionally, an
eGFR time series is interpreted by a human expert labelling it as stable or
unstable. While this approach works for individual patients, the time consuming
nature of it precludes the quick evaluation of risk in large numbers of
patients. However, automating this process poses significant challenges as eGFR
measurements are usually recorded at irregular intervals and the series of
measurements differs in length between patients. Here we present a two-tier
system to automatically classify an eGFR trend. First, we model the time series
using Gaussian process regression (GPR) to fill in `gaps' by resampling a fixed
size vector of fifty time-dependent observations. Second, we classify the
resampled eGFR time series using a K-NN/SVM classifier, and evaluate its
performance via 5-fold cross validation. Using this approach we achieved an
F-score of 0.90, compared to 0.96 for 5 human experts when scored amongst
themselves.
|
Multimodal Sparse Coding for Event Detection | Unsupervised feature learning methods have proven effective for
classification tasks based on a single modality. We present multimodal sparse
coding for learning feature representations shared across multiple modalities.
The shared representations are applied to multimedia event detection (MED) and
evaluated in comparison to unimodal counterparts, as well as other feature
learning methods such as GMM supervectors and sparse RBM. We report the
cross-validated classification accuracy and mean average precision of the MED
system trained on features learned from our unimodal and multimodal settings
for a subset of the TRECVID MED 2014 dataset.
|
On the boosting ability of top-down decision tree learning algorithm for
multiclass classification | We analyze the performance of the top-down multiclass classification
algorithm for decision tree learning called LOMtree, recently proposed in the
literature Choromanska and Langford (2014) for solving efficiently
classification problems with very large number of classes. The algorithm online
optimizes the objective function which simultaneously controls the depth of the
tree and its statistical accuracy. We prove important properties of this
objective and explore its connection to three well-known entropy-based decision
tree objectives, i.e. Shannon entropy, Gini-entropy and its modified version,
for which instead online optimization schemes were not yet developed. We show,
via boosting-type guarantees, that maximizing the considered objective leads
also to the reduction of all of these entropy-based objectives. The bounds we
obtain critically depend on the strong-concavity properties of the
entropy-based criteria, where the mildest dependence on the number of classes
(only logarithmic) corresponds to the Shannon entropy.
|
Biologically Inspired Radio Signal Feature Extraction with Sparse
Denoising Autoencoders | Automatic modulation classification (AMC) is an important task for modern
communication systems; however, it is a challenging problem when signal
features and precise models for generating each modulation may be unknown. We
present a new biologically-inspired AMC method without the need for models or
manually specified features --- thus removing the requirement for expert prior
knowledge. We accomplish this task using regularized stacked sparse denoising
autoencoders (SSDAs). Our method selects efficient classification features
directly from raw in-phase/quadrature (I/Q) radio signals in an unsupervised
manner. These features are then used to construct higher-complexity abstract
features which can be used for automatic modulation classification. We
demonstrate this process using a dataset generated with a software defined
radio, consisting of random input bits encoded in 100-sample segments of
various common digital radio modulations. Our results show correct
classification rates of > 99% at 7.5 dB signal-to-noise ratio (SNR) and > 92%
at 0 dB SNR in a 6-way classification test. Our experiments demonstrate a
dramatically new and broadly applicable mechanism for performing AMC and
related tasks without the need for expert-defined or modulation-specific signal
information.
|
Learning Convolutional Neural Networks for Graphs | Numerous important problems can be framed as learning from graph data. We
propose a framework for learning convolutional neural networks for arbitrary
graphs. These graphs may be undirected, directed, and with both discrete and
continuous node and edge attributes. Analogous to image-based convolutional
networks that operate on locally connected regions of the input, we present a
general approach to extracting locally connected regions from graphs. Using
established benchmark data sets, we demonstrate that the learned feature
representations are competitive with state of the art graph kernels and that
their computation is highly efficient.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.