title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
A Unified Neural Network Approach for Estimating Travel Time and
Distance for a Taxi Trip | In building intelligent transportation systems such as taxi or rideshare
services, accurate prediction of travel time and distance is crucial for
customer experience and resource management. Using the NYC taxi dataset, which
contains taxi trips data collected from GPS-enabled taxis [23], this paper
investigates the use of deep neural networks to jointly predict taxi trip time
and distance. We propose a model, called ST-NN (Spatio-Temporal Neural
Network), which first predicts the travel distance between an origin and a
destination GPS coordinate, then combines this prediction with the time of day
to predict the travel time. The beauty of ST-NN is that it uses only the raw
trips data without requiring further feature engineering and provides a joint
estimate of travel time and distance. We compare the performance of ST-NN to
that of state-of-the-art travel time estimation methods, and we observe that
the proposed approach generalizes better than state-of-the-art methods. We show
that ST-NN approach significantly reduces the mean absolute error for both
predicted travel time and distance, about 17% for travel time prediction. We
also observe that the proposed approach is more robust to outliers present in
the dataset by testing the performance of ST-NN on the datasets with and
without outliers.
|
Deep Learning in Multiple Multistep Time Series Prediction | The project aims to research on combining deep learning specifically
Long-Short Memory (LSTM) and basic statistics in multiple multistep time series
prediction. LSTM can dive into all the pages and learn the general trends of
variation in a large scope, while the well selected medians for each page can
keep the special seasonality of different pages so that the future trend will
not fluctuate too much from the reality. A recent Kaggle competition on 145K
Web Traffic Time Series Forecasting [1] is used to thoroughly illustrate and
test this idea.
|
Sign-Constrained Regularized Loss Minimization | In practical analysis, domain knowledge about analysis target has often been
accumulated, although, typically, such knowledge has been discarded in the
statistical analysis stage, and the statistical tool has been applied as a
black box. In this paper, we introduce sign constraints that are a handy and
simple representation for non-experts in generic learning problems. We have
developed two new optimization algorithms for the sign-constrained regularized
loss minimization, called the sign-constrained Pegasos (SC-Pega) and the
sign-constrained SDCA (SC-SDCA), by simply inserting the sign correction step
into the original Pegasos and SDCA, respectively. We present theoretical
analyses that guarantee that insertion of the sign correction step does not
degrade the convergence rate for both algorithms. Two applications, where the
sign-constrained learning is effective, are presented. The one is exploitation
of prior information about correlation between explanatory variables and a
target variable. The other is introduction of the sign-constrained to
SVM-Pairwise method. Experimental results demonstrate significant improvement
of generalization performance by introducing sign constraints in both
applications.
|
Provably Fair Representations | Machine learning systems are increasingly used to make decisions about
people's lives, such as whether to give someone a loan or whether to interview
someone for a job. This has led to considerable interest in making such machine
learning systems fair. One approach is to transform the input data used by the
algorithm. This can be achieved by passing each input data point through a
representation function prior to its use in training or testing. Techniques for
learning such representation functions from data have been successful
empirically, but typically lack theoretical fairness guarantees. We show that
it is possible to prove that a representation function is fair according to
common measures of both group and individual fairness, as well as useful with
respect to a target task. These provable properties can be used in a governance
model involving a data producer, a data user and a data regulator, where there
is a separation of concerns between fairness and target task utility to ensure
transparency and prevent perverse incentives. We formally define the 'cost of
mistrust' of using this model compared to the setting where there is a single
trusted party, and provide bounds on this cost in particular cases. We present
a practical approach to learning fair representation functions and apply it to
financial and criminal justice datasets. We evaluate the fairness and utility
of these representation functions using measures motivated by our theoretical
results.
|
Sum-Product-Quotient Networks | We present a novel tractable generative model that extends Sum-Product
Networks (SPNs) and significantly boosts their power. We call it
Sum-Product-Quotient Networks (SPQNs), whose core concept is to incorporate
conditional distributions into the model by direct computation using quotient
nodes, e.g. $P(A|B) = \frac{P(A,B)}{P(B)}$. We provide sufficient conditions
for the tractability of SPQNs that generalize and relax the decomposable and
complete tractability conditions of SPNs. These relaxed conditions give rise to
an exponential boost to the expressive efficiency of our model, i.e. we prove
that there are distributions which SPQNs can compute efficiently but require
SPNs to be of exponential size. Thus, we narrow the gap in expressivity between
tractable graphical models and other Neural Network-based generative models.
|
AMBER: Adaptive Multi-Batch Experience Replay for Continuous Action
Control | In this paper, a new adaptive multi-batch experience replay scheme is
proposed for proximal policy optimization (PPO) for continuous action control.
On the contrary to original PPO, the proposed scheme uses the batch samples of
past policies as well as the current policy for the update for the next policy,
where the number of the used past batches is adaptively determined based on the
oldness of the past batches measured by the average importance sampling (IS)
weight. The new algorithm constructed by combining PPO with the proposed
multi-batch experience replay scheme maintains the advantages of original PPO
such as random mini-batch sampling and small bias due to low IS weights by
storing the pre-computed advantages and values and adaptively determining the
mini-batch size. Numerical results show that the proposed method significantly
increases the speed and stability of convergence on various continuous control
tasks compared to original PPO.
|
Self-Taught Support Vector Machine | In this paper, a new approach for classification of target task using limited
labeled target data as well as enormous unlabeled source data is proposed which
is called self-taught learning. The target and source data can be drawn from
different distributions. In the previous approaches, covariate shift assumption
is considered where the marginal distributions p(x) change over domains and the
conditional distributions p(y|x) remain the same. In our approach, we propose a
new objective function which simultaneously learns a common space T(.) where
the conditional distributions over domains p(T(x)|y) remain the same and learns
robust SVM classifiers for target task using both source and target data in the
new representation. Hence, in the proposed objective function, the hidden label
of the source data is also incorporated. We applied the proposed approach on
Caltech-256, MSRC+LMO datasets and compared the performance of our algorithm to
the available competing methods. Our method has a superior performance to the
successful existing algorithms.
|
An Improved Naive Bayes Classifier-based Noise Detection Technique for
Classifying User Phone Call Behavior | The presence of noisy instances in mobile phone data is a fundamental issue
for classifying user phone call behavior (i.e., accept, reject, missed and
outgoing), with many potential negative consequences. The classification
accuracy may decrease and the complexity of the classifiers may increase due to
the number of redundant training samples. To detect such noisy instances from a
training dataset, researchers use naive Bayes classifier (NBC) as it identifies
misclassified instances by taking into account independence assumption and
conditional probabilities of the attributes. However, some of these
misclassified instances might indicate usages behavioral patterns of individual
mobile phone users. Existing naive Bayes classifier based noise detection
techniques have not considered this issue and, thus, are lacking in
classification accuracy. In this paper, we propose an improved noise detection
technique based on naive Bayes classifier for effectively classifying users'
phone call behaviors. In order to improve the classification accuracy, we
effectively identify noisy instances from the training dataset by analyzing the
behavioral patterns of individuals. We dynamically determine a noise threshold
according to individual's unique behavioral patterns by using both the naive
Bayes classifier and Laplace estimator. We use this noise threshold to identify
noisy instances. To measure the effectiveness of our technique in classifying
user phone call behavior, we employ the most popular classification algorithm
(e.g., decision tree). Experimental results on the real phone call log dataset
show that our proposed technique more accurately identifies the noisy instances
from the training datasets that leads to better classification accuracy.
|
Additivity of Information in Multilayer Networks via Additive Gaussian
Noise Transforms | Multilayer (or deep) networks are powerful probabilistic models based on
multiple stages of a linear transform followed by a non-linear (possibly
random) function. In general, the linear transforms are defined by matrices and
the non-linear functions are defined by information channels. These models have
gained great popularity due to their ability to characterize complex
probabilistic relationships arising in a wide variety of inference problems.
The contribution of this paper is a new method for analyzing the fundamental
limits of statistical inference in settings where the model is known. The
validity of our method can be established in a number of settings and is
conjectured to hold more generally. A key assumption made throughout is that
the matrices are drawn randomly from orthogonally invariant distributions.
Our method yields explicit formulas for 1) the mutual information; 2) the
minimum mean-squared error (MMSE); 3) the existence and locations of certain
phase-transitions with respect to the problem parameters; and 4) the stationary
points for the state evolution of approximate message passing algorithms. When
applied to the special case of models with multivariate Gaussian channels our
method is rigorous and has close connections to free probability theory for
random matrices. When applied to the general case of non-Gaussian channels, our
method provides a simple alternative to the replica method from statistical
physics. A key observation is that the combined effects of the individual
components in the model (namely the matrices and the channels) are additive
when viewed in a certain transform domain.
|
Is Epicurus the father of Reinforcement Learning? | The Epicurean Philosophy is commonly thought as simplistic and hedonistic.
Here I discuss how this is a misconception and explore its link to
Reinforcement Learning. Based on the letters of Epicurus, I construct an
objective function for hedonism which turns out to be equivalent of the
Reinforcement Learning objective function when omitting the discount factor. I
then discuss how Plato and Aristotle 's views that can be also loosely linked
to Reinforcement Learning, as well as their weaknesses in relationship to it.
Finally, I emphasise the close affinity of the Epicurean views and the Bellman
equation.
|
Towards Scalable Spectral Clustering via Spectrum-Preserving
Sparsification | The eigendeomposition of nearest-neighbor (NN) graph Laplacian matrices is
the main computational bottleneck in spectral clustering. In this work, we
introduce a highly-scalable, spectrum-preserving graph sparsification algorithm
that enables to build ultra-sparse NN (u-NN) graphs with guaranteed
preservation of the original graph spectrums, such as the first few
eigenvectors of the original graph Laplacian. Our approach can immediately lead
to scalable spectral clustering of large data networks without sacrificing
solution quality. The proposed method starts from constructing low-stretch
spanning trees (LSSTs) from the original graphs, which is followed by
iteratively recovering small portions of "spectrally critical" off-tree edges
to the LSSTs by leveraging a spectral off-tree embedding scheme. To determine
the suitable amount of off-tree edges to be recovered to the LSSTs, an
eigenvalue stability checking scheme is proposed, which enables to robustly
preserve the first few Laplacian eigenvectors within the sparsified graph.
Additionally, an incremental graph densification scheme is proposed for
identifying extra edges that have been missing in the original NN graphs but
can still play important roles in spectral clustering tasks. Our experimental
results for a variety of well-known data sets show that the proposed method can
dramatically reduce the complexity of NN graphs, leading to significant
speedups in spectral clustering.
|
Deep Imitation Learning for Complex Manipulation Tasks from Virtual
Reality Teleoperation | Imitation learning is a powerful paradigm for robot skill acquisition.
However, obtaining demonstrations suitable for learning a policy that maps from
raw pixels to actions can be challenging. In this paper we describe how
consumer-grade Virtual Reality headsets and hand tracking hardware can be used
to naturally teleoperate robots to perform complex tasks. We also describe how
imitation learning can learn deep neural network policies (mapping from pixels
to actions) that can acquire the demonstrated skills. Our experiments showcase
the effectiveness of our approach for learning visuomotor skills.
|
Social Attention: Modeling Attention in Human Crowds | Robots that navigate through human crowds need to be able to plan safe,
efficient, and human predictable trajectories. This is a particularly
challenging problem as it requires the robot to predict future human
trajectories within a crowd where everyone implicitly cooperates with each
other to avoid collisions. Previous approaches to human trajectory prediction
have modeled the interactions between humans as a function of proximity.
However, that is not necessarily true as some people in our immediate vicinity
moving in the same direction might not be as important as other people that are
further away, but that might collide with us in the future. In this work, we
propose Social Attention, a novel trajectory prediction model that captures the
relative importance of each person when navigating in the crowd, irrespective
of their proximity. We demonstrate the performance of our method against a
state-of-the-art approach on two publicly available crowd datasets and analyze
the trained attention model to gain a better understanding of which surrounding
agents humans attend to, when navigating in a crowd.
|
Hyperparameter Importance Across Datasets | With the advent of automated machine learning, automated hyperparameter
optimization methods are by now routinely used in data mining. However, this
progress is not yet matched by equal progress on automatic analyses that yield
information beyond performance-optimizing hyperparameter settings. In this
work, we aim to answer the following two questions: Given an algorithm, what
are generally its most important hyperparameters, and what are typically good
values for these? We present methodology and a framework to answer these
questions based on meta-learning across many datasets. We apply this
methodology using the experimental meta-data available on OpenML to determine
the most important hyperparameters of support vector machines, random forests
and Adaboost, and to infer priors for all their hyperparameters. The results,
obtained fully automatically, provide a quantitative basis to focus efforts in
both manual algorithm design and in automated hyperparameter optimization. The
conducted experiments confirm that the hyperparameters selected by the proposed
method are indeed the most important ones and that the obtained priors also
lead to statistically significant improvements in hyperparameter optimization.
|
On the Runtime-Efficacy Trade-off of Anomaly Detection Techniques for
Real-Time Streaming Data | Ever growing volume and velocity of data coupled with decreasing attention
span of end users underscore the critical need for real-time analytics. In this
regard, anomaly detection plays a key role as an application as well as a means
to verify data fidelity. Although the subject of anomaly detection has been
researched for over 100 years in a multitude of disciplines such as, but not
limited to, astronomy, statistics, manufacturing, econometrics, marketing, most
of the existing techniques cannot be used as is on real-time data streams.
Further, the lack of characterization of performance -- both with respect to
real-timeliness and accuracy -- on production data sets makes model selection
very challenging. To this end, we present an in-depth analysis, geared towards
real-time streaming data, of anomaly detection techniques. Given the
requirements with respect to real-timeliness and accuracy, the analysis
presented in this paper should serve as a guide for selection of the "best"
anomaly detection technique. To the best of our knowledge, this is the first
characterization of anomaly detection techniques proposed in very diverse set
of fields, using production data sets corresponding to a wide set of
application domains.
|
Bayesian Hypernetworks | We study Bayesian hypernetworks: a framework for approximate Bayesian
inference in neural networks. A Bayesian hypernetwork $\h$ is a neural network
which learns to transform a simple noise distribution, $p(\vec\epsilon) =
\N(\vec 0,\mat I)$, to a distribution $q(\pp) := q(h(\vec\epsilon))$ over the
parameters $\pp$ of another neural network (the "primary network")\@. We train
$q$ with variational inference, using an invertible $\h$ to enable efficient
estimation of the variational lower bound on the posterior $p(\pp | \D)$ via
sampling. In contrast to most methods for Bayesian deep learning, Bayesian
hypernets can represent a complex multimodal approximate posterior with
correlations between parameters, while enabling cheap iid sampling of~$q(\pp)$.
In practice, Bayesian hypernets can provide a better defense against
adversarial examples than dropout, and also exhibit competitive performance on
a suite of tasks which evaluate model uncertainty, including regularization,
active learning, and anomaly detection.
|
Sparse Weighted Canonical Correlation Analysis | Given two data matrices $X$ and $Y$, sparse canonical correlation analysis
(SCCA) is to seek two sparse canonical vectors $u$ and $v$ to maximize the
correlation between $Xu$ and $Yv$. However, classical and sparse CCA models
consider the contribution of all the samples of data matrices and thus cannot
identify an underlying specific subset of samples. To this end, we propose a
novel sparse weighted canonical correlation analysis (SWCCA), where weights are
used for regularizing different samples. We solve the $L_0$-regularized SWCCA
($L_0$-SWCCA) using an alternating iterative algorithm. We apply $L_0$-SWCCA to
synthetic data and real-world data to demonstrate its effectiveness and
superiority compared to related methods. Lastly, we consider also SWCCA with
different penalties like LASSO (Least absolute shrinkage and selection
operator) and Group LASSO, and extend it for integrating more than three data
matrices.
|
Deep Learning for Case-Based Reasoning through Prototypes: A Neural
Network that Explains Its Predictions | Deep neural networks are widely used for classification. These deep models
often suffer from a lack of interpretability -- they are particularly difficult
to understand because of their non-linear nature. As a result, neural networks
are often treated as "black box" models, and in the past, have been trained
purely to optimize the accuracy of predictions. In this work, we create a novel
network architecture for deep learning that naturally explains its own
reasoning for each prediction. This architecture contains an autoencoder and a
special prototype layer, where each unit of that layer stores a weight vector
that resembles an encoded training input. The encoder of the autoencoder allows
us to do comparisons within the latent space, while the decoder allows us to
visualize the learned prototypes. The training objective has four terms: an
accuracy term, a term that encourages every prototype to be similar to at least
one encoded input, a term that encourages every encoded input to be close to at
least one prototype, and a term that encourages faithful reconstruction by the
autoencoder. The distances computed in the prototype layer are used as part of
the classification process. Since the prototypes are learned during training,
the learned network naturally comes with explanations for each prediction, and
the explanations are loyal to what the network actually computes.
|
Deep Regression Bayesian Network and Its Applications | Deep directed generative models have attracted much attention recently due to
their generative modeling nature and powerful data representation ability. In
this paper, we review different structures of deep directed generative models
and the learning and inference algorithms associated with the structures. We
focus on a specific structure that consists of layers of Bayesian Networks due
to the property of capturing inherent and rich dependencies among latent
variables. The major difficulty of learning and inference with deep directed
models with many latent variables is the intractable inference due to the
dependencies among the latent variables and the exponential number of latent
variable configurations. Current solutions use variational methods often
through an auxiliary network to approximate the posterior probability
inference. In contrast, inference can also be performed directly without using
any auxiliary network to maximally preserve the dependencies among the latent
variables. Specifically, by exploiting the sparse representation with the
latent space, max-max instead of max-sum operation can be used to overcome the
exponential number of latent configurations. Furthermore, the max-max operation
and augmented coordinate ascent are applied to both supervised and unsupervised
learning as well as to various inference. Quantitative evaluations on benchmark
datasets of different models are given for both data representation and feature
learning tasks.
|
Recent Advances in Zero-shot Recognition | With the recent renaissance of deep convolution neural networks, encouraging
breakthroughs have been achieved on the supervised recognition tasks, where
each class has sufficient training data and fully annotated training data.
However, to scale the recognition to a large number of classes with few or now
training samples for each class remains an unsolved problem. One approach to
scaling up the recognition is to develop models capable of recognizing unseen
categories without any training instances, or zero-shot recognition/ learning.
This article provides a comprehensive review of existing zero-shot recognition
techniques covering various aspects ranging from representations of models, and
from datasets and evaluation settings. We also overview related recognition
tasks including one-shot and open set recognition which can be used as natural
extensions of zero-shot recognition when limited number of class samples become
available or when zero-shot recognition is implemented in a real-world setting.
Importantly, we highlight the limitations of existing approaches and point out
future research directions in this existing new research area.
|
Performance Comparison of Intrusion Detection Systems and Application of
Machine Learning to Snort System | This study investigates the performance of two open source intrusion
detection systems (IDSs) namely Snort and Suricata for accurately detecting the
malicious traffic on computer networks. Snort and Suricata were installed on
two different but identical computers and the performance was evaluated at 10
Gbps network speed. It was noted that Suricata could process a higher speed of
network traffic than Snort with lower packet drop rate but it consumed higher
computational resources. Snort had higher detection accuracy and was thus
selected for further experiments. It was observed that the Snort triggered a
high rate of false positive alarms. To solve this problem a Snort adaptive
plug-in was developed. To select the best performing algorithm for Snort
adaptive plug-in, an empirical study was carried out with different learning
algorithms and Support Vector Machine (SVM) was selected. A hybrid version of
SVM and Fuzzy logic produced a better detection accuracy. But the best result
was achieved using an optimised SVM with firefly algorithm with FPR (false
positive rate) as 8.6% and FNR (false negative rate) as 2.2%, which is a good
result. The novelty of this work is the performance comparison of two IDSs at
10 Gbps and the application of hybrid and optimised machine learning algorithms
to Snort.
|
Manifold regularization based on Nystr{\"o}m type subsampling | In this paper, we study the Nystr{\"o}m type subsampling for large scale
kernel methods to reduce the computational complexities of big data. We discuss
the multi-penalty regularization scheme based on Nystr{\"o}m type subsampling
which is motivated from well-studied manifold regularization schemes. We
develop a theoretical analysis of multi-penalty least-square regularization
scheme under the general source condition in vector-valued function setting,
therefore the results can also be applied to multi-task learning problems. We
achieve the optimal minimax convergence rates of multi-penalty regularization
using the concept of effective dimension for the appropriate subsampling size.
We discuss an aggregation approach based on linear function strategy to combine
various Nystr{\"o}m approximants. Finally, we demonstrate the performance of
multi-penalty regularization based on Nystr{\"o}m type subsampling on
Caltech-101 data set for multi-class image classification and NSL-KDD benchmark
data set for intrusion detection problem.
|
A Method of Generating Random Weights and Biases in Feedforward Neural
Networks with Random Hidden Nodes | Neural networks with random hidden nodes have gained increasing interest from
researchers and practical applications. This is due to their unique features
such as very fast training and universal approximation property. In these
networks the weights and biases of hidden nodes determining the nonlinear
feature mapping are set randomly and are not learned. Appropriate selection of
the intervals from which weights and biases are selected is extremely
important. This topic has not yet been sufficiently explored in the literature.
In this work a method of generating random weights and biases is proposed. This
method generates the parameters of the hidden nodes in such a way that
nonlinear fragments of the activation functions are located in the input space
regions with data and can be used to construct the surface approximating a
nonlinear target function. The weights and biases are dependent on the input
data range and activation function type. The proposed methods allows us to
control the generalization degree of the model. These all lead to improvement
in approximation performance of the network. Several experiments show very
promising results.
|
User Modelling for Avoiding Overfitting in Interactive Knowledge
Elicitation for Prediction | In human-in-the-loop machine learning, the user provides information beyond
that in the training data. Many algorithms and user interfaces have been
designed to optimize and facilitate this human--machine interaction; however,
fewer studies have addressed the potential defects the designs can cause.
Effective interaction often requires exposing the user to the training data or
its statistics. The design of the system is then critical, as this can lead to
double use of data and overfitting, if the user reinforces noisy patterns in
the data. We propose a user modelling methodology, by assuming simple rational
behaviour, to correct the problem. We show, in a user study with 48
participants, that the method improves predictive performance in a sparse
linear regression sentiment analysis task, where graded user knowledge on
feature relevance is elicited. We believe that the key idea of inferring user
knowledge with probabilistic user models has general applicability in guarding
against overfitting and improving interactive machine learning.
|
Graph Convolutional Networks for Classification with a Structured Label
Space | It is a usual practice to ignore any structural information underlying
classes in multi-class classification. In this paper, we propose a graph
convolutional network (GCN) augmented neural network classifier to exploit a
known, underlying graph structure of labels. The proposed approach resembles an
(approximate) inference procedure in, for instance, a conditional random field
(CRF). We evaluate the proposed approach on document classification and object
recognition and report both accuracies and graph-theoretic metrics that
correspond to the consistency of the model's prediction. The experiment results
reveal that the proposed model outperforms a baseline method which ignores the
graph structures of a label space in terms of graph-theoretic metrics.
|
Two-stage Algorithm for Fairness-aware Machine Learning | Algorithmic decision making process now affects many aspects of our lives.
Standard tools for machine learning, such as classification and regression, are
subject to the bias in data, and thus direct application of such off-the-shelf
tools could lead to a specific group being unfairly discriminated. Removing
sensitive attributes of data does not solve this problem because a
\textit{disparate impact} can arise when non-sensitive attributes and sensitive
attributes are correlated. Here, we study a fair machine learning algorithm
that avoids such a disparate impact when making a decision. Inspired by the
two-stage least squares method that is widely used in the field of economics,
we propose a two-stage algorithm that removes bias in the training data. The
proposed algorithm is conceptually simple. Unlike most of existing fair
algorithms that are designed for classification tasks, the proposed method is
able to (i) deal with regression tasks, (ii) combine explanatory attributes to
remove reverse discrimination, and (iii) deal with numerical sensitive
attributes. The performance and fairness of the proposed algorithm are
evaluated in simulations with synthetic and real-world datasets.
|
A Deep Incremental Boltzmann Machine for Modeling Context in Robots | Context is an essential capability for robots that are to be as adaptive as
possible in challenging environments. Although there are many context modeling
efforts, they assume a fixed structure and number of contexts. In this paper,
we propose an incremental deep model that extends Restricted Boltzmann
Machines. Our model gets one scene at a time, and gradually extends the
contextual model when necessary, either by adding a new context or a new
context layer to form a hierarchy. We show on a scene classification benchmark
that our method converges to a good estimate of the contexts of the scenes, and
performs better or on-par on several tasks compared to other incremental models
or non-incremental models.
|
CINet: A Learning Based Approach to Incremental Context Modeling in
Robots | There have been several attempts at modeling context in robots. However,
either these attempts assume a fixed number of contexts or use a rule-based
approach to determine when to increment the number of contexts. In this paper,
we pose the task of when to increment as a learning problem, which we solve
using a Recurrent Neural Network. We show that the network successfully (with
98\% testing accuracy) learns to predict when to increment, and demonstrate, in
a scene modeling problem (where the correct number of contexts is not known),
that the robot increments the number of contexts in an expected manner (i.e.,
the entropy of the system is reduced). We also present how the incremental
model can be used for various scene reasoning tasks.
|
Potential Conditional Mutual Information: Estimators, Properties and
Applications | The conditional mutual information I(X;Y|Z) measures the average information
that X and Y contain about each other given Z. This is an important primitive
in many learning problems including conditional independence testing, graphical
model inference, causal strength estimation and time-series problems. In
several applications, it is desirable to have a functional purely of the
conditional distribution p_{Y|X,Z} rather than of the joint distribution
p_{X,Y,Z}. We define the potential conditional mutual information as the
conditional mutual information calculated with a modified joint distribution
p_{Y|X,Z} q_{X,Z}, where q_{X,Z} is a potential distribution, fixed airport. We
develop K nearest neighbor based estimators for this functional, employing
importance sampling, and a coupling trick, and prove the finite k consistency
of such an estimator. We demonstrate that the estimator has excellent practical
performance and show an application in dynamical system inference.
|
Automated Scalable Bayesian Inference via Hilbert Coresets | The automation of posterior inference in Bayesian data analysis has enabled
experts and nonexperts alike to use more sophisticated models, engage in faster
exploratory modeling and analysis, and ensure experimental reproducibility.
However, standard automated posterior inference algorithms are not tractable at
the scale of massive modern datasets, and modifications to make them so are
typically model-specific, require expert tuning, and can break theoretical
guarantees on inferential quality. Building on the Bayesian coresets framework,
this work instead takes advantage of data redundancy to shrink the dataset
itself as a preprocessing step, providing fully-automated, scalable Bayesian
inference with theoretical guarantees. We begin with an intuitive reformulation
of Bayesian coreset construction as sparse vector sum approximation, and
demonstrate that its automation and performance-based shortcomings arise from
the use of the supremum norm. To address these shortcomings we develop Hilbert
coresets, i.e., Bayesian coresets constructed under a norm induced by an
inner-product on the log-likelihood function space. We propose two Hilbert
coreset construction algorithms---one based on importance sampling, and one
based on the Frank-Wolfe algorithm---along with theoretical guarantees on
approximation quality as a function of coreset size. Since the exact
computation of the proposed inner-products is model-specific, we automate the
construction with a random finite-dimensional projection of the log-likelihood
functions. The resulting automated coreset construction algorithm is simple to
implement, and experiments on a variety of models with real and synthetic
datasets show that it provides high-quality posterior approximations and a
significant reduction in the computational cost of inference.
|
A deep generative model for single-cell RNA sequencing with application
to detecting differentially expressed genes | We propose a probabilistic model for interpreting gene expression levels that
are observed through single-cell RNA sequencing. In the model, each cell has a
low-dimensional latent representation. Additional latent variables account for
technical effects that may erroneously set some observations of gene expression
levels to zero. Conditional distributions are specified by neural networks,
giving the proposed model enough flexibility to fit the data well. We use
variational inference and stochastic optimization to approximate the posterior
distribution. The inference procedure scales to over one million cells, whereas
competing algorithms do not. Even for smaller datasets, for several tasks, the
proposed procedure outperforms state-of-the-art methods like ZIFA and
ZINB-WaVE. We also extend our framework to take into account batch effects and
other confounding factors and propose a natural Bayesian hypothesis framework
for differential expression that outperforms tradition DESeq2.
|
Burn-In Demonstrations for Multi-Modal Imitation Learning | Recent work on imitation learning has generated policies that reproduce
expert behavior from multi-modal data. However, past approaches have focused
only on recreating a small number of distinct, expert maneuvers, or have relied
on supervised learning techniques that produce unstable policies. This work
extends InfoGAIL, an algorithm for multi-modal imitation learning, to reproduce
behavior over an extended period of time. Our approach involves reformulating
the typical imitation learning setting to include "burn-in demonstrations" upon
which policies are conditioned at test time. We demonstrate that our approach
outperforms standard InfoGAIL in maximizing the mutual information between
predicted and unseen style labels in road scene simulations, and we show that
our method leads to policies that imitate expert autonomous driving systems
over long time horizons.
|
A simple data discretizer | Data discretization is an important step in the process of machine learning,
since it is easier for classifiers to deal with discrete attributes rather than
continuous attributes. Over the years, several methods of performing
discretization such as Boolean Reasoning, Equal Frequency Binning, Entropy have
been proposed, explored, and implemented. In this article, a simple supervised
discretization approach is introduced. The prime goal of MIL is to maximize
classification accuracy of classifier, minimizing loss of information while
discretization of continuous attributes. The performance of the suggested
approach is compared with the supervised discretization algorithm Minimum
Information Loss (MIL), using the state-of-the-art rule inductive algorithms-
J48 (Java implementation of C4.5 classifier). The presented approach is,
indeed, the modified version of MIL. The empirical results show that the
modified approach performs better in several cases in comparison to the
original MIL algorithm and Minimum Description Length Principle (MDLP) .
|
Dropout as a Low-Rank Regularizer for Matrix Factorization | Regularization for matrix factorization (MF) and approximation problems has
been carried out in many different ways. Due to its popularity in deep
learning, dropout has been applied also for this class of problems. Despite its
solid empirical performance, the theoretical properties of dropout as a
regularizer remain quite elusive for this class of problems. In this paper, we
present a theoretical analysis of dropout for MF, where Bernoulli random
variables are used to drop columns of the factors. We demonstrate the
equivalence between dropout and a fully deterministic model for MF in which the
factors are regularized by the sum of the product of squared Euclidean norms of
the columns. Additionally, we inspect the case of a variable sized
factorization and we prove that dropout achieves the global minimum of a convex
approximation problem with (squared) nuclear norm regularization. As a result,
we conclude that dropout can be used as a low-rank regularizer with data
dependent singular-value thresholding.
|
CM-GANs: Cross-modal Generative Adversarial Networks for Common
Representation Learning | It is known that the inconsistent distribution and representation of
different modalities, such as image and text, cause the heterogeneity gap that
makes it challenging to correlate such heterogeneous data. Generative
adversarial networks (GANs) have shown its strong ability of modeling data
distribution and learning discriminative representation, existing GANs-based
works mainly focus on generative problem to generate new data. We have
different goal, aim to correlate heterogeneous data, by utilizing the power of
GANs to model cross-modal joint distribution. Thus, we propose Cross-modal GANs
to learn discriminative common representation for bridging heterogeneity gap.
The main contributions are: (1) Cross-modal GANs architecture is proposed to
model joint distribution over data of different modalities. The inter-modality
and intra-modality correlation can be explored simultaneously in generative and
discriminative models. Both of them beat each other to promote cross-modal
correlation learning. (2) Cross-modal convolutional autoencoders with
weight-sharing constraint are proposed to form generative model. They can not
only exploit cross-modal correlation for learning common representation, but
also preserve reconstruction information for capturing semantic consistency
within each modality. (3) Cross-modal adversarial mechanism is proposed, which
utilizes two kinds of discriminative models to simultaneously conduct
intra-modality and inter-modality discrimination. They can mutually boost to
make common representation more discriminative by adversarial training process.
To the best of our knowledge, our proposed CM-GANs approach is the first to
utilize GANs to perform cross-modal common representation learning. Experiments
are conducted to verify the performance of our proposed approach on cross-modal
retrieval paradigm, compared with 10 methods on 3 cross-modal datasets.
|
Subsampling for Ridge Regression via Regularized Volume Sampling | Given $n$ vectors $\mathbf{x}_i\in \mathbb{R}^d$, we want to fit a linear
regression model for noisy labels $y_i\in\mathbb{R}$. The ridge estimator is a
classical solution to this problem. However, when labels are expensive, we are
forced to select only a small subset of vectors $\mathbf{x}_i$ for which we
obtain the labels $y_i$. We propose a new procedure for selecting the subset of
vectors, such that the ridge estimator obtained from that subset offers strong
statistical guarantees in terms of the mean squared prediction error over the
entire dataset of $n$ labeled vectors. The number of labels needed is
proportional to the statistical dimension of the problem which is often much
smaller than $d$. Our method is an extension of a joint subsampling procedure
called volume sampling. A second major contribution is that we speed up volume
sampling so that it is essentially as efficient as leverage scores, which is
the main i.i.d. subsampling procedure for this task. Finally, we show
theoretically and experimentally that volume sampling has a clear advantage
over any i.i.d. sampling when labels are expensive.
|
Parametric t-Distributed Stochastic Exemplar-centered Embedding | Parametric embedding methods such as parametric t-SNE (pt-SNE) have been
widely adopted for data visualization and out-of-sample data embedding without
further computationally expensive optimization or approximation. However, the
performance of pt-SNE is highly sensitive to the hyper-parameter batch size due
to conflicting optimization goals, and often produces dramatically different
embeddings with different choices of user-defined perplexities. To effectively
solve these issues, we present parametric t-distributed stochastic
exemplar-centered embedding methods. Our strategy learns embedding parameters
by comparing given data only with precomputed exemplars, resulting in a cost
function with linear computational and memory complexity, which is further
reduced by noise contrastive samples. Moreover, we propose a shallow embedding
network with high-order feature interactions for data visualization, which is
much easier to tune but produces comparable performance in contrast to a deep
neural network employed by pt-SNE. We empirically demonstrate, using several
benchmark datasets, that our proposed methods significantly outperform pt-SNE
in terms of robustness, visual effects, and quantitative evaluations.
|
When Point Process Meets RNNs: Predicting Fine-Grained User Interests
with Mutual Behavioral Infectivity | Predicting fine-grained interests of users with temporal behavior is
important to personalization and information filtering applications. However,
existing interest prediction methods are incapable of capturing the subtle
degreed user interests towards particular items, and the internal time-varying
drifting attention of individuals is not studied yet. Moreover, the prediction
process can also be affected by inter-personal influence, known as behavioral
mutual infectivity. Inspired by point process in modeling temporal point
process, in this paper we present a deep prediction method based on two
recurrent neural networks (RNNs) to jointly model each user's continuous
browsing history and asynchronous event sequences in the context of inter-user
behavioral mutual infectivity. Our model is able to predict the fine-grained
interest from a user regarding a particular item and corresponding timestamps
when an occurrence of event takes place. The proposed approach is more flexible
to capture the dynamic characteristic of event sequences by using the temporal
point process to model event data and timely update its intensity function by
RNNs. Furthermore, to improve the interpretability of the model, the attention
mechanism is introduced to emphasize both intra-personal and inter-personal
behavior influence over time. Experiments on real datasets demonstrate that our
model outperforms the state-of-the-art methods in fine-grained user interest
prediction.
|
Regularizing Deep Neural Networks by Noise: Its Interpretation and
Optimization | Overfitting is one of the most critical challenges in deep neural networks,
and there are various types of regularization methods to improve generalization
performance. Injecting noises to hidden units during training, e.g., dropout,
is known as a successful regularizer, but it is still not clear enough why such
training techniques work well in practice and how we can maximize their benefit
in the presence of two conflicting objectives---optimizing to true data
distribution and preventing overfitting by regularization. This paper addresses
the above issues by 1) interpreting that the conventional training methods with
regularization by noise injection optimize the lower bound of the true
objective and 2) proposing a technique to achieve a tighter lower bound using
multiple noise samples per training example in a stochastic gradient descent
iteration. We demonstrate the effectiveness of our idea in several computer
vision applications.
|
Community Aware Random Walk for Network Embedding | Social network analysis provides meaningful information about behavior of
network members that can be used for diverse applications such as
classification, link prediction. However, network analysis is computationally
expensive because of feature learning for different applications. In recent
years, many researches have focused on feature learning methods in social
networks. Network embedding represents the network in a lower dimensional
representation space with the same properties which presents a compressed
representation of the network. In this paper, we introduce a novel algorithm
named "CARE" for network embedding that can be used for different types of
networks including weighted, directed and complex. Current methods try to
preserve local neighborhood information of nodes, whereas the proposed method
utilizes local neighborhood and community information of network nodes to cover
both local and global structure of social networks. CARE builds customized
paths, which are consisted of local and global structure of network nodes, as a
basis for network embedding and uses the Skip-gram model to learn
representation vector of nodes. Subsequently, stochastic gradient descent is
applied to optimize our objective function and learn the final representation
of nodes. Our method can be scalable when new nodes are appended to network
without information loss. Parallelize generation of customized random walks is
also used for speeding up CARE. We evaluate the performance of CARE on multi
label classification and link prediction tasks. Experimental results on various
networks indicate that the proposed method outperforms others in both Micro and
Macro-f1 measures for different size of training data.
|
Near-optimal Sample Complexity Bounds for Robust Learning of Gaussians
Mixtures via Compression Schemes | We prove that $\tilde{\Theta}(k d^2 / \varepsilon^2)$ samples are necessary
and sufficient for learning a mixture of $k$ Gaussians in $\mathbb{R}^d$, up to
error $\varepsilon$ in total variation distance. This improves both the known
upper bounds and lower bounds for this problem. For mixtures of axis-aligned
Gaussians, we show that $\tilde{O}(k d / \varepsilon^2)$ samples suffice,
matching a known lower bound. Moreover, these results hold in the
agnostic-learning/robust-estimation setting as well, where the target
distribution is only approximately a mixture of Gaussians.
The upper bound is shown using a novel technique for distribution learning
based on a notion of `compression.' Any class of distributions that allows such
a compression scheme can also be learned with few samples. Moreover, if a class
of distributions has such a compression scheme, then so do the classes of
products and mixtures of those distributions. The core of our main result is
showing that the class of Gaussians in $\mathbb{R}^d$ admits a small-sized
compression scheme.
|
Mental Sampling in Multimodal Representations | Both resources in the natural environment and concepts in a semantic space
are distributed "patchily", with large gaps in between the patches. To describe
people's internal and external foraging behavior, various random walk models
have been proposed. In particular, internal foraging has been modeled as
sampling: in order to gather relevant information for making a decision, people
draw samples from a mental representation using random-walk algorithms such as
Markov chain Monte Carlo (MCMC). However, two common empirical observations
argue against simple sampling algorithms such as MCMC. First, the spatial
structure is often best described by a L\'evy flight distribution: the
probability of the distance between two successive locations follows a
power-law on the distances. Second, the temporal structure of the sampling that
humans and other animals produce have long-range, slowly decaying serial
correlations characterized as $1/f$-like fluctuations. We propose that mental
sampling is not done by simple MCMC, but is instead adapted to multimodal
representations and is implemented by Metropolis-coupled Markov chain Monte
Carlo (MC$^3$), one of the first algorithms developed for sampling from
multimodal distributions. MC$^3$ involves running multiple Markov chains in
parallel but with target distributions of different temperatures, and it swaps
the states of the chains whenever a better location is found. Heated chains
more readily traverse valleys in the probability landscape to propose moves to
far-away peaks, while the colder chains make the local steps that explore the
current peak or patch. We show that MC$^3$ generates distances between
successive samples that follow a L\'evy flight distribution and $1/f$-like
serial correlations, providing a single mechanistic account of these two
puzzling empirical phenomena.
|
Learners that Use Little Information | We study learning algorithms that are restricted to using a small amount of
information from their input sample. We introduce a category of learning
algorithms we term $d$-bit information learners, which are algorithms whose
output conveys at most $d$ bits of information of their input. A central theme
in this work is that such algorithms generalize.
We focus on the learning capacity of these algorithms, and prove sample
complexity bounds with tight dependencies on the confidence and error
parameters. We also observe connections with well studied notions such as
sample compression schemes, Occam's razor, PAC-Bayes and differential privacy.
We discuss an approach that allows us to prove upper bounds on the amount of
information that algorithms reveal about their inputs, and also provide a lower
bound by showing a simple concept class for which every (possibly randomized)
empirical risk minimizer must reveal a lot of information. On the other hand,
we show that in the distribution-dependent setting every VC class has empirical
risk minimizers that do not reveal a lot of information.
|
Robust Decentralized Learning Using ADMM with Unreliable Agents | Many machine learning problems can be formulated as consensus optimization
problems which can be solved efficiently via a cooperative multi-agent system.
However, the agents in the system can be unreliable due to a variety of
reasons: noise, faults and attacks. Providing erroneous updates leads the
optimization process in a wrong direction, and degrades the performance of
distributed machine learning algorithms. This paper considers the problem of
decentralized learning using ADMM in the presence of unreliable agents. First,
we rigorously analyze the effect of erroneous updates (in ADMM learning
iterations) on the convergence behavior of multi-agent system. We show that the
algorithm linearly converges to a neighborhood of the optimal solution under
certain conditions and characterize the neighborhood size analytically. Next,
we provide guidelines for network design to achieve a faster convergence. We
also provide conditions on the erroneous updates for exact convergence to the
optimal solution. Finally, to mitigate the influence of unreliable agents, we
propose \textsf{ROAD}, a robust variant of ADMM, and show its resilience to
unreliable agents with an exact convergence to the optimum.
|
Self-Supervised Visual Planning with Temporal Skip Connections | In order to autonomously learn wide repertoires of complex skills, robots
must be able to learn from their own autonomously collected data, without human
supervision. One learning signal that is always available for autonomously
collected data is prediction: if a robot can learn to predict the future, it
can use this predictive model to take actions to produce desired outcomes, such
as moving an object to a particular location. However, in complex open-world
scenarios, designing a representation for prediction is difficult. In this
work, we instead aim to enable self-supervised robotic learning through direct
video prediction: instead of attempting to design a good representation, we
directly predict what the robot will see next, and then use this model to
achieve desired goals. A key challenge in video prediction for robotic
manipulation is handling complex spatial arrangements such as occlusions. To
that end, we introduce a video prediction model that can keep track of objects
through occlusion by incorporating temporal skip-connections. Together with a
novel planning criterion and action space formulation, we demonstrate that this
model substantially outperforms prior work on video prediction-based control.
Our results show manipulation of objects not seen during training, handling
multiple objects, and pushing objects around obstructions. These results
represent a significant advance in the range and complexity of skills that can
be performed entirely with self-supervised robotic learning.
|
Learning Infinite RBMs with Frank-Wolfe | In this work, we propose an infinite restricted Boltzmann machine~(RBM),
whose maximum likelihood estimation~(MLE) corresponds to a constrained convex
optimization. We consider the Frank-Wolfe algorithm to solve the program, which
provides a sparse solution that can be interpreted as inserting a hidden unit
at each iteration, so that the optimization process takes the form of a
sequence of finite models of increasing complexity. As a side benefit, this can
be used to easily and efficiently identify an appropriate number of hidden
units during the optimization. The resulting model can also be used as an
initialization for typical state-of-the-art RBM training algorithms such as
contrastive divergence, leading to models with consistently higher test
likelihood than random initialization.
|
Facial Keypoints Detection | Detect facial keypoints is a critical element in face recognition. However,
there is difficulty to catch keypoints on the face due to complex influences
from original images, and there is no guidance to suitable algorithms. In this
paper, we study different algorithms that can be applied to locate keyponits.
Specifically: our framework (1)prepare the data for further investigation
(2)Using PCA and LBP to process the data (3) Apply different algorithms to
analysis data, including linear regression models, tree based model, neural
network and convolutional neural network, etc. Finally we will give our
conclusion and further research topic. A comprehensive set of experiments on
dataset demonstrates the effectiveness of our framework.
|
CNNComparator: Comparative Analytics of Convolutional Neural Networks | Convolutional neural networks (CNNs) are widely used in many image
recognition tasks due to their extraordinary performance. However, training a
good CNN model can still be a challenging task. In a training process, a CNN
model typically learns a large number of parameters over time, which usually
results in different performance. Often, it is difficult to explore the
relationships between the learned parameters and the model performance due to a
large number of parameters and different random initializations. In this paper,
we present a visual analytics approach to compare two different snapshots of a
trained CNN model taken after different numbers of epochs, so as to provide
some insight into the design or the training of a better CNN model. Our system
compares snapshots by exploring the differences in operation parameters and the
corresponding blob data at different levels. A case study has been conducted to
demonstrate the effectiveness of our system.
|
Text2Action: Generative Adversarial Synthesis from Language to Action | In this paper, we propose a generative model which learns the relationship
between language and human action in order to generate a human action sequence
given a sentence describing human behavior. The proposed generative model is a
generative adversarial network (GAN), which is based on the sequence to
sequence (SEQ2SEQ) model. Using the proposed generative network, we can
synthesize various actions for a robot or a virtual agent using a text encoder
recurrent neural network (RNN) and an action decoder RNN. The proposed
generative network is trained from 29,770 pairs of actions and sentence
annotations extracted from MSR-Video-to-Text (MSR-VTT), a large-scale video
dataset. We demonstrate that the network can generate human-like actions which
can be transferred to a Baxter robot, such that the robot performs an action
based on a provided sentence. Results show that the proposed generative network
correctly models the relationship between language and action and can generate
a diverse set of actions from the same sentence.
|
Accelerated Block Coordinate Proximal Gradients with Applications in
High Dimensional Statistics | Nonconvex optimization problems arise in different research fields and arouse
lots of attention in signal processing, statistics and machine learning. In
this work, we explore the accelerated proximal gradient method and some of its
variants which have been shown to converge under nonconvex context recently. We
show that a novel variant proposed here, which exploits adaptive momentum and
block coordinate update with specific update rules, further improves the
performance of a broad class of nonconvex problems. In applications to sparse
linear regression with regularizations like Lasso, grouped Lasso, capped
$\ell_1$ and SCAP, the proposed scheme enjoys provable local linear
convergence, with experimental justification.
|
Information-Theoretic Representation Learning for Positive-Unlabeled
Classification | Recent advances in weakly supervised classification allow us to train a
classifier only from positive and unlabeled (PU) data. However, existing PU
classification methods typically require an accurate estimate of the
class-prior probability, which is a critical bottleneck particularly for
high-dimensional data. This problem has been commonly addressed by applying
principal component analysis in advance, but such unsupervised dimension
reduction can collapse underlying class structure. In this paper, we propose a
novel representation learning method from PU data based on the
information-maximization principle. Our method does not require class-prior
estimation and thus can be used as a preprocessing method for PU
classification. Through experiments, we demonstrate that our method combined
with deep neural networks highly improves the accuracy of PU class-prior
estimation, leading to state-of-the-art PU classification performance.
|
Robust Locally-Linear Controllable Embedding | Embed-to-control (E2C) is a model for solving high-dimensional optimal
control problems by combining variational auto-encoders with locally-optimal
controllers. However, the E2C model suffers from two major drawbacks: 1) its
objective function does not correspond to the likelihood of the data sequence
and 2) the variational encoder used for embedding typically has large
variational approximation error, especially when there is noise in the system
dynamics. In this paper, we present a new model for learning robust
locally-linear controllable embedding (RCE). Our model directly estimates the
predictive conditional density of the future observation given the current one,
while introducing the bottleneck between the current and future observations.
Although the bottleneck provides a natural embedding candidate for control, our
RCE model introduces additional specific structures in the generative graphical
model so that the model dynamics can be robustly linearized. We also propose a
principled variational approximation of the embedding posterior that takes the
future observation into account, and thus, makes the variational approximation
more robust against the noise. Experimental results show that RCE outperforms
the E2C model, and does so significantly when the underlying dynamics is noisy.
|
A systematic study of the class imbalance problem in convolutional
neural networks | In this study, we systematically investigate the impact of class imbalance on
classification performance of convolutional neural networks (CNNs) and compare
frequently used methods to address the issue. Class imbalance is a common
problem that has been comprehensively studied in classical machine learning,
yet very limited systematic research is available in the context of deep
learning. In our study, we use three benchmark datasets of increasing
complexity, MNIST, CIFAR-10 and ImageNet, to investigate the effects of
imbalance on classification and perform an extensive comparison of several
methods to address the issue: oversampling, undersampling, two-phase training,
and thresholding that compensates for prior class probabilities. Our main
evaluation metric is area under the receiver operating characteristic curve
(ROC AUC) adjusted to multi-class tasks since overall accuracy metric is
associated with notable difficulties in the context of imbalanced data. Based
on results from our experiments we conclude that (i) the effect of class
imbalance on classification performance is detrimental; (ii) the method of
addressing class imbalance that emerged as dominant in almost all analyzed
scenarios was oversampling; (iii) oversampling should be applied to the level
that completely eliminates the imbalance, whereas the optimal undersampling
ratio depends on the extent of imbalance; (iv) as opposed to some classical
machine learning models, oversampling does not cause overfitting of CNNs; (v)
thresholding should be applied to compensate for prior class probabilities when
overall number of properly classified cases is of interest.
|
The Scaling Limit of High-Dimensional Online Independent Component
Analysis | We analyze the dynamics of an online algorithm for independent component
analysis in the high-dimensional scaling limit. As the ambient dimension tends
to infinity, and with proper time scaling, we show that the time-varying joint
empirical measure of the target feature vector and the estimates provided by
the algorithm will converge weakly to a deterministic measured-valued process
that can be characterized as the unique solution of a nonlinear PDE. Numerical
solutions of this PDE, which involves two spatial variables and one time
variable, can be efficiently obtained. These solutions provide detailed
information about the performance of the ICA algorithm, as many practical
performance metrics are functionals of the joint empirical measures. Numerical
simulations show that our asymptotic analysis is accurate even for moderate
dimensions. In addition to providing a tool for understanding the performance
of the algorithm, our PDE analysis also provides useful insight. In particular,
in the high-dimensional limit, the original coupled dynamics associated with
the algorithm will be asymptotically "decoupled", with each coordinate
independently solving a 1-D effective minimization problem via stochastic
gradient descent. Exploiting this insight to design new algorithms for
achieving optimal trade-offs between computational and statistical efficiency
may prove an interesting line of future research.
|
Manifold Regularization for Kernelized LSTD | Policy evaluation or value function or Q-function approximation is a key
procedure in reinforcement learning (RL). It is a necessary component of policy
iteration and can be used for variance reduction in policy gradient methods.
Therefore its quality has a significant impact on most RL algorithms. Motivated
by manifold regularized learning, we propose a novel kernelized policy
evaluation method that takes advantage of the intrinsic geometry of the state
space learned from data, in order to achieve better sample efficiency and
higher accuracy in Q-function approximation. Applying the proposed method in
the Least-Squares Policy Iteration (LSPI) framework, we observe superior
performance compared to widely used parametric basis functions on two standard
benchmarks in terms of policy quality.
|
NeuralPower: Predict and Deploy Energy-Efficient Convolutional Neural
Networks | "How much energy is consumed for an inference made by a convolutional neural
network (CNN)?" With the increased popularity of CNNs deployed on the
wide-spectrum of platforms (from mobile devices to workstations), the answer to
this question has drawn significant attention. From lengthening battery life of
mobile devices to reducing the energy bill of a datacenter, it is important to
understand the energy efficiency of CNNs during serving for making an
inference, before actually training the model. In this work, we propose
NeuralPower: a layer-wise predictive framework based on sparse polynomial
regression, for predicting the serving energy consumption of a CNN deployed on
any GPU platform. Given the architecture of a CNN, NeuralPower provides an
accurate prediction and breakdown for power and runtime across all layers in
the whole network, helping machine learners quickly identify the power,
runtime, or energy bottlenecks. We also propose the "energy-precision ratio"
(EPR) metric to guide machine learners in selecting an energy-efficient CNN
architecture that better trades off the energy consumption and prediction
accuracy. The experimental results show that the prediction accuracy of the
proposed NeuralPower outperforms the best published model to date, yielding an
improvement in accuracy of up to 68.5%. We also assess the accuracy of
predictions at the network level, by predicting the runtime, power, and energy
of state-of-the-art CNN architectures, achieving an average accuracy of 88.24%
in runtime, 88.34% in power, and 97.21% in energy. We comprehensively
corroborate the effectiveness of NeuralPower as a powerful framework for
machine learners by testing it on different GPU platforms and Deep Learning
software tools.
|
A General Framework for Robust Interactive Learning | We propose a general framework for interactively learning models, such as
(binary or non-binary) classifiers, orderings/rankings of items, or clusterings
of data points. Our framework is based on a generalization of Angluin's
equivalence query model and Littlestone's online learning model: in each
iteration, the algorithm proposes a model, and the user either accepts it or
reveals a specific mistake in the proposal. The feedback is correct only with
probability $p > 1/2$ (and adversarially incorrect with probability $1 - p$),
i.e., the algorithm must be able to learn in the presence of arbitrary noise.
The algorithm's goal is to learn the ground truth model using few iterations.
Our general framework is based on a graph representation of the models and
user feedback. To be able to learn efficiently, it is sufficient that there be
a graph $G$ whose nodes are the models and (weighted) edges capture the user
feedback, with the property that if $s, s^*$ are the proposed and target
models, respectively, then any (correct) user feedback $s'$ must lie on a
shortest $s$-$s^*$ path in $G$. Under this one assumption, there is a natural
algorithm reminiscent of the Multiplicative Weights Update algorithm, which
will efficiently learn $s^*$ even in the presence of noise in the user's
feedback.
From this general result, we rederive with barely any extra effort classic
results on learning of classifiers and a recent result on interactive
clustering; in addition, we easily obtain new interactive learning algorithms
for ordering/ranking.
|
Generalization in Deep Learning | This paper provides theoretical insights into why and how deep learning can
generalize well, despite its large capacity, complexity, possible algorithmic
instability, nonrobustness, and sharp minima, responding to an open question in
the literature. We also discuss approaches to provide non-vacuous
generalization guarantees for deep learning. Based on theoretical observations,
we propose new open problems and discuss the limitations of our results.
|
Safe Learning of Quadrotor Dynamics Using Barrier Certificates | To effectively control complex dynamical systems, accurate nonlinear models
are typically needed. However, these models are not always known. In this
paper, we present a data-driven approach based on Gaussian processes that
learns models of quadrotors operating in partially unknown environments. What
makes this challenging is that if the learning process is not carefully
controlled, the system will go unstable, i.e., the quadcopter will crash. To
this end, barrier certificates are employed for safe learning. The barrier
certificates establish a non-conservative forward invariant safe region, in
which high probability safety guarantees are provided based on the statistics
of the Gaussian Process. A learning controller is designed to efficiently
explore those uncertain states and expand the barrier certified safe region
based on an adaptive sampling scheme. In addition, a recursive Gaussian Process
prediction method is developed to learn the complex quadrotor dynamics in
real-time. Simulation results are provided to demonstrate the effectiveness of
the proposed approach.
|
Calibrated Boosting-Forest | Excellent ranking power along with well calibrated probability estimates are
needed in many classification tasks. In this paper, we introduce a technique,
Calibrated Boosting-Forest that captures both. This novel technique is an
ensemble of gradient boosting machines that can support both continuous and
binary labels. While offering superior ranking power over any individual
regression or classification model, Calibrated Boosting-Forest is able to
preserve well calibrated posterior probabilities. Along with these benefits, we
provide an alternative to the tedious step of tuning gradient boosting
machines. We demonstrate that tuning Calibrated Boosting-Forest can be reduced
to a simple hyper-parameter selection. We further establish that increasing
this hyper-parameter improves the ranking performance under a diminishing
return. We examine the effectiveness of Calibrated Boosting-Forest on
ligand-based virtual screening where both continuous and binary labels are
available and compare the performance of Calibrated Boosting-Forest with
logistic regression, gradient boosting machine and deep learning. Calibrated
Boosting-Forest achieved an approximately 48% improvement compared to a
state-of-art deep learning model. Moreover, it achieved around 95% improvement
on probability quality measurement compared to the best individual gradient
boosting machine. Calibrated Boosting-Forest offers a benchmark demonstration
that in the field of ligand-based virtual screening, deep learning is not the
universally dominant machine learning model and good calibrated probabilities
can better facilitate virtual screening process.
|
A Geometric View of Optimal Transportation and Generative Model | In this work, we show the intrinsic relations between optimal transportation
and convex geometry, especially the variational approach to solve Alexandrov
problem: constructing a convex polytope with prescribed face normals and
volumes. This leads to a geometric interpretation to generative models, and
leads to a novel framework for generative models. By using the optimal
transportation view of GAN model, we show that the discriminator computes the
Kantorovich potential, the generator calculates the transportation map. For a
large class of transportation costs, the Kantorovich potential can give the
optimal transportation map by a close-form formula. Therefore, it is sufficient
to solely optimize the discriminator. This shows the adversarial competition
can be avoided, and the computational architecture can be simplified.
Preliminary experimental results show the geometric method outperforms WGAN for
approximating probability measures with multiple clusters in low dimensional
space.
|
The Feeling of Success: Does Touch Sensing Help Predict Grasp Outcomes? | A successful grasp requires careful balancing of the contact forces. Deducing
whether a particular grasp will be successful from indirect measurements, such
as vision, is therefore quite challenging, and direct sensing of contacts
through touch sensing provides an appealing avenue toward more successful and
consistent robotic grasping. However, in order to fully evaluate the value of
touch sensing for grasp outcome prediction, we must understand how touch
sensing can influence outcome prediction accuracy when combined with other
modalities. Doing so using conventional model-based techniques is exceptionally
difficult. In this work, we investigate the question of whether touch sensing
aids in predicting grasp outcomes within a multimodal sensing framework that
combines vision and touch. To that end, we collected more than 9,000 grasping
trials using a two-finger gripper equipped with GelSight high-resolution
tactile sensors on each finger, and evaluated visuo-tactile deep neural network
models to directly predict grasp outcomes from either modality individually,
and from both modalities together. Our experimental results indicate that
incorporating tactile readings substantially improve grasping performance.
|
Entanglement Entropy of Target Functions for Image Classification and
Convolutional Neural Network | The success of deep convolutional neural network (CNN) in computer vision
especially image classification problems requests a new information theory for
function of image, instead of image itself. In this article, after establishing
a deep mathematical connection between image classification problem and quantum
spin model, we propose to use entanglement entropy, a generalization of
classical Boltzmann-Shannon entropy, as a powerful tool to characterize the
information needed for representation of general function of image. We prove
that there is a sub-volume-law bound for entanglement entropy of target
functions of reasonable image classification problems. Therefore target
functions of image classification only occupy a small subspace of the whole
Hilbert space. As a result, a neural network with polynomial number of
parameters is efficient for representation of such target functions of image.
The concept of entanglement entropy can also be useful to characterize the
expressive power of different neural networks. For example, we show that to
maintain the same expressive power, number of channels $D$ in a convolutional
neural network should scale with the number of convolution layers $n_c$ as
$D\sim D_0^{\frac{1}{n_c}}$. Therefore, deeper CNN with large $n_c$ is more
efficient than shallow ones.
|
Is Simple Better? Revisiting Non-linear Matrix Factorization for
Learning Incomplete Ratings | Matrix factorization techniques have been widely used as a method for
collaborative filtering for recommender systems. In recent times, different
variants of deep learning algorithms have been explored in this setting to
improve the task of making a personalized recommendation with user-item
interaction data. The idea that the mapping between the latent user or item
factors and the original features is highly nonlinear suggest that classical
matrix factorization techniques are no longer sufficient. In this paper, we
propose a multilayer nonlinear semi-nonnegative matrix factorization method,
with the motivation that user-item interactions can be modeled more accurately
using a linear combination of non-linear item features. Firstly, we learn
latent factors for representations of users and items from the designed
multilayer nonlinear Semi-NMF approach using explicit ratings. Secondly, the
architecture built is compared with deep-learning algorithms like Restricted
Boltzmann Machine and state-of-the-art Deep Matrix factorization techniques. By
using both supervised rate prediction task and unsupervised clustering in
latent item space, we demonstrate that our proposed approach achieves better
generalization ability in prediction as well as comparable representation
ability as deep matrix factorization in the clustering task.
|
Large Scale Graph Learning from Smooth Signals | Graphs are a prevalent tool in data science, as they model the inherent
structure of the data. They have been used successfully in unsupervised and
semi-supervised learning. Typically they are constructed either by connecting
nearest samples, or by learning them from data, solving an optimization
problem. While graph learning does achieve a better quality, it also comes with
a higher computational cost. In particular, the current state-of-the-art model
cost is $\mathcal{O}(n^2)$ for $n$ samples. In this paper, we show how to scale
it, obtaining an approximation with leading cost of $\mathcal{O}(n\log(n))$,
with quality that approaches the exact graph learning model. Our algorithm uses
known approximate nearest neighbor techniques to reduce the number of
variables, and automatically selects the correct parameters of the model,
requiring a single intuitive input: the desired edge density.
|
Deep Self-Paced Learning for Person Re-Identification | Person re-identification (Re-ID) usually suffers from noisy samples with
background clutter and mutual occlusion, which makes it extremely difficult to
distinguish different individuals across the disjoint camera views. In this
paper, we propose a novel deep self-paced learning (DSPL) algorithm to
alleviate this problem, in which we apply a self-paced constraint and symmetric
regularization to help the relative distance metric training the deep neural
network, so as to learn the stable and discriminative features for person
Re-ID. Firstly, we propose a soft polynomial regularizer term which can derive
the adaptive weights to samples based on both the training loss and model age.
As a result, the high-confidence fidelity samples will be emphasized and the
low-confidence noisy samples will be suppressed at early stage of the whole
training process. Such a learning regime is naturally implemented under a
self-paced learning (SPL) framework, in which samples weights are adaptively
updated based on both model age and sample loss using an alternative
optimization method. Secondly, we introduce a symmetric regularizer term to
revise the asymmetric gradient back-propagation derived by the relative
distance metric, so as to simultaneously minimize the intra-class distance and
maximize the inter-class distance in each triplet unit. Finally, we build a
part-based deep neural network, in which the features of different body parts
are first discriminately learned in the lower convolutional layers and then
fused in the higher fully connected layers. Experiments on several benchmark
datasets have demonstrated the superior performance of our method as compared
with the state-of-the-art approaches.
|
Lung Cancer Screening Using Adaptive Memory-Augmented Recurrent Networks | In this paper, we investigate the effectiveness of deep learning techniques
for lung nodule classification in computed tomography scans. Using less than
10,000 training examples, our deep networks perform two times better than a
standard radiology software. Visualization of the networks' neurons reveals
semantically meaningful features that are consistent with the clinical
knowledge and radiologists' perception. Our paper also proposes a novel
framework for rapidly adapting deep networks to the radiologists' feedback, or
change in the data due to the shift in sensor's resolution or patient
population. The classification accuracy of our approach remains above 80% while
popular deep networks' accuracy is around chance. Finally, we provide in-depth
analysis of our framework by asking a radiologist to examine important
networks' features and perform blind re-labeling of networks' mistakes.
|
On the Hardness of Inventory Management with Censored Demand Data | We consider a repeated newsvendor problem where the inventory manager has no
prior information about the demand, and can access only censored/sales data. In
analogy to multi-armed bandit problems, the manager needs to simultaneously
"explore" and "exploit" with her inventory decisions, in order to minimize the
cumulative cost. We make no probabilistic assumptions---importantly,
independence or time stationarity---regarding the mechanism that creates the
demand sequence. Our goal is to shed light on the hardness of the problem, and
to develop policies that perform well with respect to the regret criterion,
that is, the difference between the cumulative cost of a policy and that of the
best fixed action/static inventory decision in hindsight, uniformly over all
feasible demand sequences. We show that a simple randomized policy, termed the
Exponentially Weighted Forecaster, combined with a carefully designed cost
estimator, achieves optimal scaling of the expected regret (up to logarithmic
factors) with respect to all three key primitives: the number of time periods,
the number of inventory decisions available, and the demand support. Through
this result, we derive an important insight: the benefit from "information
stalking" as well as the cost of censoring are both negligible in this dynamic
learning problem, at least with respect to the regret criterion. Furthermore,
we modify the proposed policy in order to perform well in terms of the tracking
regret, that is, using as benchmark the best sequence of inventory decisions
that switches a limited number of times. Numerical experiments suggest that the
proposed approach outperforms existing ones (that are tailored to, or
facilitated by, time stationarity) on nonstationary demand models. Finally, we
extend the proposed approach and its analysis to a "combinatorial" version of
the repeated newsvendor problem.
|
A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised
Learning | This paper takes a step towards temporal reasoning in a dynamically changing
video, not in the pixel space that constitutes its frames, but in a latent
space that describes the non-linear dynamics of the objects in its world. We
introduce the Kalman variational auto-encoder, a framework for unsupervised
learning of sequential data that disentangles two latent representations: an
object's representation, coming from a recognition model, and a latent state
describing its dynamics. As a result, the evolution of the world can be
imagined and missing data imputed, both without the need to generate high
dimensional frames at each time step. The model is trained end-to-end on videos
of a variety of simulated physical systems, and outperforms competing methods
in generative and missing data imputation tasks.
|
TensorQuant - A Simulation Toolbox for Deep Neural Network Quantization | Recent research implies that training and inference of deep neural networks
(DNN) can be computed with low precision numerical representations of the
training/test data, weights and gradients without a general loss in accuracy.
The benefit of such compact representations is twofold: they allow a
significant reduction of the communication bottleneck in distributed DNN
training and faster neural network implementations on hardware accelerators
like FPGAs. Several quantization methods have been proposed to map the original
32-bit floating point problem to low-bit representations. While most related
publications validate the proposed approach on a single DNN topology, it
appears to be evident, that the optimal choice of the quantization method and
number of coding bits is topology dependent. To this end, there is no general
theory available, which would allow users to derive the optimal quantization
during the design of a DNN topology. In this paper, we present a quantization
tool box for the TensorFlow framework. TensorQuant allows a transparent
quantization simulation of existing DNN topologies during training and
inference. TensorQuant supports generic quantization methods and allows
experimental evaluation of the impact of the quantization on single layers as
well as on the full topology. In a first series of experiments with
TensorQuant, we show an analysis of fix-point quantizations of popular CNN
topologies.
|
Discriminative Learning of Prediction Intervals | In this work we consider the task of constructing prediction intervals in an
inductive batch setting. We present a discriminative learning framework which
optimizes the expected error rate under a budget constraint on the interval
sizes. Most current methods for constructing prediction intervals offer
guarantees for a single new test point. Applying these methods to multiple test
points can result in a high computational overhead and degraded statistical
guarantees. By focusing on expected errors, our method allows for variability
in the per-example conditional error rates. As we demonstrate both analytically
and empirically, this flexibility can increase the overall accuracy, or
alternatively, reduce the average interval size.
While the problem we consider is of a regressive flavor, the loss we use is
combinatorial. This allows us to provide PAC-style, finite-sample guarantees.
Computationally, we show that our original objective is NP-hard, and suggest a
tractable convex surrogate. We conclude with a series of experimental
evaluations.
|
Spectral Algorithms for Computing Fair Support Vector Machines | Classifiers and rating scores are prone to implicitly codifying biases, which
may be present in the training data, against protected classes (i.e., age,
gender, or race). So it is important to understand how to design classifiers
and scores that prevent discrimination in predictions. This paper develops
computationally tractable algorithms for designing accurate but fair support
vector machines (SVM's). Our approach imposes a constraint on the covariance
matrices conditioned on each protected class, which leads to a nonconvex
quadratic constraint in the SVM formulation. We develop iterative algorithms to
compute fair linear and kernel SVM's, which solve a sequence of relaxations
constructed using a spectral decomposition of the nonconvex constraint. Its
effectiveness in achieving high prediction accuracy while ensuring fairness is
shown through numerical experiments on several data sets.
|
Searching for Activation Functions | The choice of activation functions in deep networks has a significant effect
on the training dynamics and task performance. Currently, the most successful
and widely-used activation function is the Rectified Linear Unit (ReLU).
Although various hand-designed alternatives to ReLU have been proposed, none
have managed to replace it due to inconsistent gains. In this work, we propose
to leverage automatic search techniques to discover new activation functions.
Using a combination of exhaustive and reinforcement learning-based search, we
discover multiple novel activation functions. We verify the effectiveness of
the searches by conducting an empirical evaluation with the best discovered
activation function. Our experiments show that the best discovered activation
function, $f(x) = x \cdot \text{sigmoid}(\beta x)$, which we name Swish, tends
to work better than ReLU on deeper models across a number of challenging
datasets. For example, simply replacing ReLUs with Swish units improves top-1
classification accuracy on ImageNet by 0.9\% for Mobile NASNet-A and 0.6\% for
Inception-ResNet-v2. The simplicity of Swish and its similarity to ReLU make it
easy for practitioners to replace ReLUs with Swish units in any neural network.
|
Gradient-free Policy Architecture Search and Adaptation | We develop a method for policy architecture search and adaptation via
gradient-free optimization which can learn to perform autonomous driving tasks.
By learning from both demonstration and environmental reward we develop a model
that can learn with relatively few early catastrophic failures. We first learn
an architecture of appropriate complexity to perceive aspects of world state
relevant to the expert demonstration, and then mitigate the effect of
domain-shift during deployment by adapting a policy demonstrated in a source
domain to rewards obtained in a target environment. We show that our approach
allows safer learning than baseline methods, offering a reduced cumulative
crash metric over the agent's lifetime as it learns to drive in a realistic
simulated environment.
|
Pushing the envelope in deep visual recognition for mobile platforms | Image classification is the task of assigning to an input image a label from
a fixed set of categories. One of its most important applicative fields is that
of robotics, in particular the needing of a robot to be aware of what's around
and the consequent exploitation of that information as a benefit for its tasks.
In this work we consider the problem of a robot that enters a new environment
and wants to understand visual data coming from its camera, so to extract
knowledge from them. As main novelty we want to overcome the needing of a
physical robot, as it could be expensive and unhandy, so to hopefully enhance,
speed up and ease the research in this field. That's why we propose to develop
an application for a mobile platform that wraps several deep visual recognition
tasks. First we deal with a simple Image classification, testing a model
obtained from an AlexNet trained on the ILSVRC 2012 dataset. Several photo
settings are considered to better understand which factors affect most the
quality of classification. For the same purpose we are interested to integrate
the classification task with an extra module dealing with segmentation of the
object inside the image. In particular we propose a technique for extracting
the object shape and moving out all the background, so to focus the
classification only on the region occupied by the object. Another significant
task that is included is that of object discovery. Its purpose is to simulate
the situation in which the robot needs a certain object to complete one of its
activities. It starts searching for what it needs by looking around and trying
to understand the location of the object by scanning the surrounding
environment. Finally we provide a tool for dealing with the creation of
customized task-specific databases, meant to better suit to one's needing in a
particular vision task.
|
Stochastic Variance Reduction for Policy Gradient Estimation | Recent advances in policy gradient methods and deep learning have
demonstrated their applicability for complex reinforcement learning problems.
However, the variance of the performance gradient estimates obtained from the
simulation is often excessive, leading to poor sample efficiency. In this
paper, we apply the stochastic variance reduced gradient descent (SVRG) to
model-free policy gradient to significantly improve the sample-efficiency. The
SVRG estimation is incorporated into a trust-region Newton conjugate gradient
framework for the policy optimization. On several Mujoco tasks, our method
achieves significantly better performance compared to the state-of-the-art
model-free policy gradient methods in robotic continuous control such as trust
region policy optimization (TRPO)
|
Boosting Adversarial Attacks with Momentum | Deep neural networks are vulnerable to adversarial examples, which poses
security concerns on these algorithms due to the potentially severe
consequences. Adversarial attacks serve as an important surrogate to evaluate
the robustness of deep learning models before they are deployed. However, most
of existing adversarial attacks can only fool a black-box model with a low
success rate. To address this issue, we propose a broad class of momentum-based
iterative algorithms to boost adversarial attacks. By integrating the momentum
term into the iterative process for attacks, our methods can stabilize update
directions and escape from poor local maxima during the iterations, resulting
in more transferable adversarial examples. To further improve the success rates
for black-box attacks, we apply momentum iterative algorithms to an ensemble of
models, and show that the adversarially trained models with a strong defense
ability are also vulnerable to our black-box attacks. We hope that the proposed
methods will serve as a benchmark for evaluating the robustness of various deep
models and defense methods. With this method, we won the first places in NIPS
2017 Non-targeted Adversarial Attack and Targeted Adversarial Attack
competitions.
|
On the challenges of learning with inference networks on sparse,
high-dimensional data | We study parameter estimation in Nonlinear Factor Analysis (NFA) where the
generative model is parameterized by a deep neural network. Recent work has
focused on learning such models using inference (or recognition) networks; we
identify a crucial problem when modeling large, sparse, high-dimensional
datasets -- underfitting. We study the extent of underfitting, highlighting
that its severity increases with the sparsity of the data. We propose methods
to tackle it via iterative optimization inspired by stochastic variational
inference \citep{hoffman2013stochastic} and improvements in the sparse data
representation used for inference. The proposed techniques drastically improve
the ability of these powerful models to fit sparse data, achieving
state-of-the-art results on a benchmark text-count dataset and excellent
results on the task of top-N recommendation.
|
Spontaneous Symmetry Breaking in Neural Networks | We propose a framework to understand the unprecedented performance and
robustness of deep neural networks using field theory. Correlations between the
weights within the same layer can be described by symmetries in that layer, and
networks generalize better if such symmetries are broken to reduce the
redundancies of the weights. Using a two parameter field theory, we find that
the network can break such symmetries itself towards the end of training in a
process commonly known in physics as spontaneous symmetry breaking. This
corresponds to a network generalizing itself without any user input layers to
break the symmetry, but by communication with adjacent layers. In the layer
decoupling limit applicable to residual networks (He et al., 2015), we show
that the remnant symmetries that survive the non-linear layers are
spontaneously broken. The Lagrangian for the non-linear and weight layers
together has striking similarities with the one in quantum field theory of a
scalar. Using results from quantum field theory we show that our framework is
able to explain many experimentally observed phenomena,such as training on
random labels with zero error (Zhang et al., 2017), the information bottleneck,
the phase transition out of it and gradient variance explosion (Shwartz-Ziv &
Tishby, 2017), shattered gradients (Balduzzi et al., 2017), and many more.
|
Primal-Dual $\pi$ Learning: Sample Complexity and Sublinear Run Time for
Ergodic Markov Decision Problems | Consider the problem of approximating the optimal policy of a Markov decision
process (MDP) by sampling state transitions. In contrast to existing
reinforcement learning methods that are based on successive approximations to
the nonlinear Bellman equation, we propose a Primal-Dual $\pi$ Learning method
in light of the linear duality between the value and policy. The $\pi$ learning
method is model-free and makes primal-dual updates to the policy and value
vectors as new data are revealed. For infinite-horizon undiscounted Markov
decision process with finite state space $S$ and finite action space $A$, the
$\pi$ learning method finds an $\epsilon$-optimal policy using the following
number of sample transitions $$ \tilde{O}( \frac{(\tau\cdot t^*_{mix})^2 |S|
|A| }{\epsilon^2} ),$$ where $t^*_{mix}$ is an upper bound of mixing times
across all policies and $\tau$ is a parameter characterizing the range of
stationary distributions across policies. The $\pi$ learning method also
applies to the computational problem of MDP where the transition probabilities
and rewards are explicitly given as the input. In the case where each state
transition can be sampled in $\tilde{O}(1)$ time, the $\pi$ learning method
gives a sublinear-time algorithm for solving the averaged-reward MDP.
|
Map-based Multi-Policy Reinforcement Learning: Enhancing Adaptability of
Robots by Deep Reinforcement Learning | In order for robots to perform mission-critical tasks, it is essential that
they are able to quickly adapt to changes in their environment as well as to
injuries and or other bodily changes. Deep reinforcement learning has been
shown to be successful in training robot control policies for operation in
complex environments. However, existing methods typically employ only a single
policy. This can limit the adaptability since a large environmental
modification might require a completely different behavior compared to the
learning environment. To solve this problem, we propose Map-based Multi-Policy
Reinforcement Learning (MMPRL), which aims to search and store multiple
policies that encode different behavioral features while maximizing the
expected reward in advance of the environment change. Thanks to these policies,
which are stored into a multi-dimensional discrete map according to its
behavioral feature, adaptation can be performed within reasonable time without
retraining the robot. An appropriate pre-trained policy from the map can be
recalled using Bayesian optimization. Our experiments show that MMPRL enables
robots to quickly adapt to large changes without requiring any prior knowledge
on the type of injuries that could occur. A highlight of the learned behaviors
can be found here: https://youtu.be/QwInbilXNOE .
|
Convolutional Recurrent Neural Networks for Electrocardiogram
Classification | We propose two deep neural network architectures for classification of
arbitrary-length electrocardiogram (ECG) recordings and evaluate them on the
atrial fibrillation (AF) classification data set provided by the PhysioNet/CinC
Challenge 2017. The first architecture is a deep convolutional neural network
(CNN) with averaging-based feature aggregation across time. The second
architecture combines convolutional layers for feature extraction with
long-short term memory (LSTM) layers for temporal aggregation of features. As a
key ingredient of our training procedure we introduce a simple data
augmentation scheme for ECG data and demonstrate its effectiveness in the AF
classification task at hand. The second architecture was found to outperform
the first one, obtaining an $F_1$ score of $82.1$% on the hidden challenge
testing set.
|
Operational thermal load forecasting in district heating networks using
machine learning and expert advice | Forecasting thermal load is a key component for the majority of optimization
solutions for controlling district heating and cooling systems. Recent studies
have analysed the results of a number of data-driven methods applied to thermal
load forecasting, this paper presents the results of combining a collection of
these individual methods in an expert system. The expert system will combine
multiple thermal load forecasts in a way that it always tracks the best expert
in the system. This solution is tested and validated using a thermal load
dataset of 27 months obtained from 10 residential buildings located in Rottne,
Sweden together with outdoor temperature information received from a weather
forecast service. The expert system is composed of the following data-driven
methods: linear regression, extremely randomized trees regression, feed-forward
neural network and support vector machine. The results of the proposed solution
are compared with the results of the individual methods.
|
Cross-Language Learning for Program Classification using Bilateral
Tree-Based Convolutional Neural Networks | Towards the vision of translating code that implements an algorithm from one
programming language into another, this paper proposes an approach for
automated program classification using bilateral tree-based convolutional
neural networks (BiTBCNNs). It is layered on top of two tree-based
convolutional neural networks (TBCNNs), each of which recognizes the algorithm
of code written in an individual programming language. The combination layer of
the networks recognizes the similarities and differences among code in
different programming languages. The BiTBCNNs are trained using the source code
in different languages but known to implement the same algorithms and/or
functionalities. For a preliminary evaluation, we use 3591 Java and 3534 C++
code snippets from 6 algorithms we crawled systematically from GitHub. We
obtained over 90% accuracy in the cross-language binary classification task to
tell whether any given two code snippets implement the same algorithm. Also,
for the algorithm classification task, i.e., to predict which one of the six
algorithm labels is implemented by an arbitrary C++ code snippet, we achieved
over 80% precision.
|
Distill-and-Compare: Auditing Black-Box Models Using Transparent Model
Distillation | Black-box risk scoring models permeate our lives, yet are typically
proprietary or opaque. We propose Distill-and-Compare, a model distillation and
comparison approach to audit such models. To gain insight into black-box
models, we treat them as teachers, training transparent student models to mimic
the risk scores assigned by black-box models. We compare the student model
trained with distillation to a second un-distilled transparent model trained on
ground-truth outcomes, and use differences between the two models to gain
insight into the black-box model. Our approach can be applied in a realistic
setting, without probing the black-box model API. We demonstrate the approach
on four public data sets: COMPAS, Stop-and-Frisk, Chicago Police, and Lending
Club. We also propose a statistical test to determine if a data set is missing
key features used to train the black-box model. Our test finds that the
ProPublica data is likely missing key feature(s) used in COMPAS.
|
Deep Gaussian Covariance Network | The correlation length-scale next to the noise variance are the most used
hyperparameters for the Gaussian processes. Typically, stationary covariance
functions are used, which are only dependent on the distances between input
points and thus invariant to the translations in the input space. The
optimization of the hyperparameters is commonly done by maximizing the log
marginal likelihood. This works quite well, if the distances are uniform
distributed. In the case of a locally adapted or even sparse input space, the
prediction of a test point can be worse dependent of its position. A possible
solution to this, is the usage of a non-stationary covariance function, where
the hyperparameters are calculated by a deep neural network. So that the
correlation length scales and possibly the noise variance are dependent on the
test point. Furthermore, different types of covariance functions are trained
simultaneously, so that the Gaussian process prediction is an additive overlay
of different covariance matrices. The right covariance functions combination
and its hyperparameters are learned by the deep neural network. Additional, the
Gaussian process will be able to be trained by batches or online and so it can
handle arbitrarily large data sets. We call this framework Deep Gaussian
Covariance Network (DGCP). There are also further extensions to this framework
possible, for example sequentially dependent problems like time series or the
local mixture of experts. The basic framework and some extension possibilities
will be presented in this work. Moreover, a comparison to some recent state of
the art surrogate model methods will be performed, also for a time dependent
problem.
|
Learning to Warm-Start Bayesian Hyperparameter Optimization | Hyperparameter optimization aims to find the optimal hyperparameter
configuration of a machine learning model, which provides the best performance
on a validation dataset. Manual search usually leads to get stuck in a local
hyperparameter configuration, and heavily depends on human intuition and
experience. A simple alternative of manual search is random/grid search on a
space of hyperparameters, which still undergoes extensive evaluations of
validation errors in order to find its best configuration. Bayesian
optimization that is a global optimization method for black-box functions is
now popular for hyperparameter optimization, since it greatly reduces the
number of validation error evaluations required, compared to random/grid
search. Bayesian optimization generally finds the best hyperparameter
configuration from random initialization without any prior knowledge. This
motivates us to let Bayesian optimization start from the configurations that
were successful on similar datasets, which are able to remarkably minimize the
number of evaluations. In this paper, we propose deep metric learning to learn
meta-features over datasets such that the similarity over them is effectively
measured by Euclidean distance between their associated meta-features. To this
end, we introduce a Siamese network composed of deep feature and meta-feature
extractors, where deep feature extractor provides a semantic representation of
each instance in a dataset and meta-feature extractor aggregates a set of deep
features to encode a single representation over a dataset. Then, our learned
meta-features are used to select a few datasets similar to the new dataset, so
that hyperparameters in similar datasets are adopted as initializations to
warm-start Bayesian hyperparameter optimization.
|
Combinatorial Penalties: Which structures are preserved by convex
relaxations? | We consider the homogeneous and the non-homogeneous convex relaxations for
combinatorial penalty functions defined on support sets. Our study identifies
key differences in the tightness of the resulting relaxations through the
notion of the lower combinatorial envelope of a set-function along with new
necessary conditions for support identification. We then propose a general
adaptive estimator for convex monotone regularizers, and derive new sufficient
conditions for support recovery in the asymptotic setting.
|
Smooth and Sparse Optimal Transport | Entropic regularization is quickly emerging as a new standard in optimal
transport (OT). It enables to cast the OT computation as a differentiable and
unconstrained convex optimization problem, which can be efficiently solved
using the Sinkhorn algorithm. However, entropy keeps the transportation plan
strictly positive and therefore completely dense, unlike unregularized OT. This
lack of sparsity can be problematic in applications where the transportation
plan itself is of interest. In this paper, we explore regularizing the primal
and dual OT formulations with a strongly convex term, which corresponds to
relaxing the dual and primal constraints with smooth approximations. We show
how to incorporate squared $2$-norm and group lasso regularizations within that
framework, leading to sparse and group-sparse transportation plans. On the
theoretical side, we bound the approximation error introduced by regularizing
the primal and dual formulations. Our results suggest that, for the regularized
primal, the approximation error can often be smaller with squared $2$-norm than
with entropic regularization. We showcase our proposed framework on the task of
color transfer.
|
Beat by Beat: Classifying Cardiac Arrhythmias with Recurrent Neural
Networks | With tens of thousands of electrocardiogram (ECG) records processed by mobile
cardiac event recorders every day, heart rhythm classification algorithms are
an important tool for the continuous monitoring of patients at risk. We utilise
an annotated dataset of 12,186 single-lead ECG recordings to build a diverse
ensemble of recurrent neural networks (RNNs) that is able to distinguish
between normal sinus rhythms, atrial fibrillation, other types of arrhythmia
and signals that are too noisy to interpret. In order to ease learning over the
temporal dimension, we introduce a novel task formulation that harnesses the
natural segmentation of ECG signals into heartbeats to drastically reduce the
number of time steps per sequence. Additionally, we extend our RNNs with an
attention mechanism that enables us to reason about which heartbeats our RNNs
focus on to make their decisions. Through the use of attention, our model
maintains a high degree of interpretability, while also achieving
state-of-the-art classification performance with an average F1 score of 0.79 on
an unseen test set (n=3,658).
|
Convergence diagnostics for stochastic gradient descent with constant
step size | Many iterative procedures in stochastic optimization exhibit a transient
phase followed by a stationary phase. During the transient phase the procedure
converges towards a region of interest, and during the stationary phase the
procedure oscillates in that region, commonly around a single point. In this
paper, we develop a statistical diagnostic test to detect such phase transition
in the context of stochastic gradient descent with constant learning rate. We
present theory and experiments suggesting that the region where the proposed
diagnostic is activated coincides with the convergence region. For a class of
loss functions, we derive a closed-form solution describing such region.
Finally, we suggest an application to speed up convergence of stochastic
gradient descent by halving the learning rate each time stationarity is
detected. This leads to a new variant of stochastic gradient descent, which in
many settings is comparable to state-of-art.
|
Fishing for Clickbaits in Social Images and Texts with
Linguistically-Infused Neural Network Models | This paper presents the results and conclusions of our participation in the
Clickbait Challenge 2017 on automatic clickbait detection in social media. We
first describe linguistically-infused neural network models and identify
informative representations to predict the level of clickbaiting present in
Twitter posts. Our models allow to answer the question not only whether a post
is a clickbait or not, but to what extent it is a clickbait post e.g., not at
all, slightly, considerably, or heavily clickbaity using a score ranging from 0
to 1. We evaluate the predictive power of models trained on varied text and
image representations extracted from tweets. Our best performing model that
relies on the tweet text and linguistic markers of biased language extracted
from the tweet and the corresponding page yields mean squared error (MSE) of
0.04, mean absolute error (MAE) of 0.16 and R2 of 0.43 on the held-out test
data. For the binary classification setup (clickbait vs. non-clickbait), our
model achieved F1 score of 0.69. We have not found that image representations
combined with text yield significant performance improvement yet. Nevertheless,
this work is the first to present preliminary analysis of objects extracted
using Google Tensorflow object detection API from images in clickbait vs.
non-clickbait Twitter posts. Finally, we outline several steps to improve model
performance as a part of the future work.
|
Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from
Simulation | Learning-based approaches to robotic manipulation are limited by the
scalability of data collection and accessibility of labels. In this paper, we
present a multi-task domain adaptation framework for instance grasping in
cluttered scenes by utilizing simulated robot experiments. Our neural network
takes monocular RGB images and the instance segmentation mask of a specified
target object as inputs, and predicts the probability of successfully grasping
the specified object for each candidate motor command. The proposed transfer
learning framework trains a model for instance grasping in simulation and uses
a domain-adversarial loss to transfer the trained model to real robots using
indiscriminate grasping data, which is available both in simulation and the
real world. We evaluate our model in real-world robot experiments, comparing it
with alternative model architectures as well as an indiscriminate grasping
baseline.
|
Domain Randomization and Generative Models for Robotic Grasping | Deep learning-based robotic grasping has made significant progress thanks to
algorithmic improvements and increased data availability. However,
state-of-the-art models are often trained on as few as hundreds or thousands of
unique object instances, and as a result generalization can be a challenge.
In this work, we explore a novel data generation pipeline for training a deep
neural network to perform grasp planning that applies the idea of domain
randomization to object synthesis. We generate millions of unique, unrealistic
procedurally generated objects, and train a deep neural network to perform
grasp planning on these objects.
Since the distribution of successful grasps for a given object can be highly
multimodal, we propose an autoregressive grasp planning model that maps sensor
inputs of a scene to a probability distribution over possible grasps. This
model allows us to sample grasps efficiently at test time (or avoid sampling
entirely).
We evaluate our model architecture and data generation pipeline in simulation
and the real world. We find we can achieve a $>$90% success rate on previously
unseen realistic objects at test time in simulation despite having only been
trained on random objects. We also demonstrate an 80% success rate on
real-world grasp attempts despite having only been trained on random simulated
objects.
|
A Bayesian Perspective on Generalization and Stochastic Gradient Descent | We consider two questions at the heart of machine learning; how can we
predict if a minimum will generalize to the test set, and why does stochastic
gradient descent find minima that generalize well? Our work responds to Zhang
et al. (2016), who showed deep neural networks can easily memorize randomly
labeled training data, despite generalizing well on real labels of the same
inputs. We show that the same phenomenon occurs in small linear models. These
observations are explained by the Bayesian evidence, which penalizes sharp
minima but is invariant to model parameterization. We also demonstrate that,
when one holds the learning rate fixed, there is an optimum batch size which
maximizes the test set accuracy. We propose that the noise introduced by small
mini-batches drives the parameters towards minima whose evidence is large.
Interpreting stochastic gradient descent as a stochastic differential equation,
we identify the "noise scale" $g = \epsilon (\frac{N}{B} - 1) \approx \epsilon
N/B$, where $\epsilon$ is the learning rate, $N$ the training set size and $B$
the batch size. Consequently the optimum batch size is proportional to both the
learning rate and the size of the training set, $B_{opt} \propto \epsilon N$.
We verify these predictions empirically.
|
S-Isomap++: Multi Manifold Learning from Streaming Data | Manifold learning based methods have been widely used for non-linear
dimensionality reduction (NLDR). However, in many practical settings, the need
to process streaming data is a challenge for such methods, owing to the high
computational complexity involved. Moreover, most methods operate under the
assumption that the input data is sampled from a single manifold, embedded in a
high dimensional space. We propose a method for streaming NLDR when the
observed data is either sampled from multiple manifolds or irregularly sampled
from a single manifold. We show that existing NLDR methods, such as Isomap,
fail in such situations, primarily because they rely on smoothness and
continuity of the underlying manifold, which is violated in the scenarios
explored in this paper. However, the proposed algorithm is able to learn
effectively in presence of multiple, and potentially intersecting, manifolds,
while allowing for the input data to arrive as a massive stream.
|
Learning Inverse Statics Models Efficiently | Online Goal Babbling and Direction Sampling are recently proposed methods for
direct learning of inverse kinematics mappings from scratch even in
high-dimensional sensorimotor spaces following the paradigm of "learning while
behaving". To learn inverse statics mappings - primarily for gravity
compensation - from scratch and without using any closed-loop controller, we
modify and enhance the Online Goal Babbling and Direction Sampling schemes.
Moreover, we exploit symmetries in the inverse statics mappings to drastically
reduce the number of samples required for learning inverse statics models.
Results for a 2R planar robot, a 3R simplified human arm, and a 4R humanoid
robot arm clearly demonstrate that their inverse statics mappings can be
learned successfully with our modified online Goal Babbling scheme.
Furthermore, we show that the number of samples required for the 2R and 3R arms
can be reduced by a factor of at least 8 and 16 resp. -depending on the number
of discovered symmetries.
|
Coded Fourier Transform | We consider the problem of computing the Fourier transform of
high-dimensional vectors, distributedly over a cluster of machines consisting
of a master node and multiple worker nodes, where the worker nodes can only
store and process a fraction of the inputs. We show that by exploiting the
algebraic structure of the Fourier transform operation and leveraging concepts
from coding theory, one can efficiently deal with the straggler effects. In
particular, we propose a computation strategy, named as coded FFT, which
achieves the optimal recovery threshold, defined as the minimum number of
workers that the master node needs to wait for in order to compute the output.
This is the first code that achieves the optimum robustness in terms of
tolerating stragglers or failures for computing Fourier transforms.
Furthermore, the reconstruction process for coded FFT can be mapped to MDS
decoding, which can be solved efficiently. Moreover, we extend coded FFT to
settings including computing general $n$-dimensional Fourier transforms, and
provide the optimal computing strategy for those settings.
|
Robust importance-weighted cross-validation under sample selection bias | Cross-validation under sample selection bias can, in principle, be done by
importance-weighting the empirical risk. However, the importance-weighted risk
estimator produces sub-optimal hyperparameter estimates in problem settings
where large weights arise with high probability. We study its sampling variance
as a function of the training data distribution and introduce a control variate
to increase its robustness to problematically large weights.
|
LASAGNE: Locality And Structure Aware Graph Node Embedding | In this work we propose Lasagne, a methodology to learn locality and
structure aware graph node embeddings in an unsupervised way. In particular, we
show that the performance of existing random-walk based approaches depends
strongly on the structural properties of the graph, e.g., the size of the
graph, whether the graph has a flat or upward-sloping Network Community Profile
(NCP), whether the graph is expander-like, whether the classes of interest are
more k-core-like or more peripheral, etc. For larger graphs with flat NCPs that
are strongly expander-like, existing methods lead to random walks that expand
rapidly, touching many dissimilar nodes, thereby leading to lower-quality
vector representations that are less useful for downstream tasks. Rather than
relying on global random walks or neighbors within fixed hop distances, Lasagne
exploits strongly local Approximate Personalized PageRank stationary
distributions to more precisely engineer local information into node
embeddings. This leads, in particular, to more meaningful and more useful
vector representations of nodes in poorly-structured graphs. We show that
Lasagne leads to significant improvement in downstream multi-label
classification for larger graphs with flat NCPs, that it is comparable for
smaller graphs with upward-sloping NCPs, and that is comparable to existing
methods for link prediction tasks.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.