title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Efficient Embedding of MPI Collectives in MXNET DAGs for scaling Deep
Learning | Availability of high performance computing infrastructures such as clusters
of GPUs and CPUs have fueled the growth of distributed learning systems. Deep
Learning frameworks express neural nets as DAGs and execute these DAGs on
computation resources such as GPUs. In this paper, we propose efficient designs
of embedding MPI collective operations into data parallel DAGs. Incorrect
designs can easily lead to deadlocks or program crashes. In particular, we
demonstrate three designs: Funneled, Concurrent communication and Dependency
chaining of using MPI collectives with DAGs. These designs automatically enable
overlap of computation with communication by allowing for concurrent execution
with the other tasks. We directly implement these designs into the KVStore API
of the MXNET. This allows us to directly leverage the rest of the
infrastructure. Using ImageNet and CIFAR data sets, we show the potential of
our designs. In particular, our designs scale to 256 GPUs with as low as 50
seconds of epoch times for ImageNet 1K datasets.
|
Neural Network Ensembles to Real-time Identification of Plug-level
Appliance Measurements | The problem of identifying end-use electrical appliances from their
individual consumption profiles, known as the appliance identification problem,
is a primary stage in both Non-Intrusive Load Monitoring (NILM) and automated
plug-wise metering. Therefore, appliance identification has received dedicated
studies with various electric appliance signatures, classification models, and
evaluation datasets. In this paper, we propose a neural network ensembles
approach to address this problem using high resolution measurements. The models
are trained on the raw current and voltage waveforms, and thus, eliminating the
need for well engineered appliance signatures. We evaluate the proposed model
on a publicly available appliance dataset from 55 residential buildings, 11
appliance categories, and over 1000 measurements. We further study the
stability of the trained models with respect to training dataset, sampling
frequency, and variations in the steady-state operation of appliances.
|
Constant Regret, Generalized Mixability, and Mirror Descent | We consider the setting of prediction with expert advice; a learner makes
predictions by aggregating those of a group of experts. Under this setting, and
for the right choice of loss function and "mixing" algorithm, it is possible
for the learner to achieve a constant regret regardless of the number of
prediction rounds. For example, a constant regret can be achieved for
\emph{mixable} losses using the \emph{aggregating algorithm}. The
\emph{Generalized Aggregating Algorithm} (GAA) is a name for a family of
algorithms parameterized by convex functions on simplices (entropies), which
reduce to the aggregating algorithm when using the \emph{Shannon entropy}
$\operatorname{S}$. For a given entropy $\Phi$, losses for which a constant
regret is possible using the \textsc{GAA} are called $\Phi$-mixable. Which
losses are $\Phi$-mixable was previously left as an open question. We fully
characterize $\Phi$-mixability and answer other open questions posed by
\cite{Reid2015}. We show that the Shannon entropy $\operatorname{S}$ is
fundamental in nature when it comes to mixability; any $\Phi$-mixable loss is
necessarily $\operatorname{S}$-mixable, and the lowest worst-case regret of the
\textsc{GAA} is achieved using the Shannon entropy. Finally, by leveraging the
connection between the \emph{mirror descent algorithm} and the update step of
the GAA, we suggest a new \emph{adaptive} generalized aggregating algorithm and
analyze its performance in terms of the regret bound.
|
Fitting New Speakers Based on a Short Untranscribed Sample | Learning-based Text To Speech systems have the potential to generalize from
one speaker to the next and thus require a relatively short sample of any new
voice. However, this promise is currently largely unrealized. We present a
method that is designed to capture a new speaker from a short untranscribed
audio sample. This is done by employing an additional network that given an
audio sample, places the speaker in the embedding space. This network is
trained as part of the speech synthesis system using various consistency
losses. Our results demonstrate a greatly improved performance on both the
dataset speakers, and, more importantly, when fitting new voices, even from
very short samples.
|
Traffic Graph Convolutional Recurrent Neural Network: A Deep Learning
Framework for Network-Scale Traffic Learning and Forecasting | Traffic forecasting is a particularly challenging application of
spatiotemporal forecasting, due to the time-varying traffic patterns and the
complicated spatial dependencies on road networks. To address this challenge,
we learn the traffic network as a graph and propose a novel deep learning
framework, Traffic Graph Convolutional Long Short-Term Memory Neural Network
(TGC-LSTM), to learn the interactions between roadways in the traffic network
and forecast the network-wide traffic state. We define the traffic graph
convolution based on the physical network topology. The relationship between
the proposed traffic graph convolution and the spectral graph convolution is
also discussed. An L1-norm on graph convolution weights and an L2-norm on graph
convolution features are added to the model's loss function to enhance the
interpretability of the proposed model. Experimental results show that the
proposed model outperforms baseline methods on two real-world traffic state
datasets. The visualization of the graph convolution weights indicates that the
proposed framework can recognize the most influential road segments in
real-world traffic networks.
|
Segmentation hi\'erarchique faiblement supervis\'ee | Image segmentation is the process of partitioning an image into a set of
meaningful regions according to some criteria. Hierarchical segmentation has
emerged as a major trend in this regard as it favors the emergence of important
regions at different scales. On the other hand, many methods allow us to have
prior information on the position of structures of interest in the images. In
this paper, we present a versatile hierarchical segmentation method that takes
into account any prior spatial information and outputs a hierarchical
segmentation that emphasizes the contours or regions of interest while
preserving the important structures in the image. An application of this method
to the weakly-supervised segmentation problem is presented.
|
A Flexible and Adaptive Framework for Abstention Under Class Imbalance | In practical applications of machine learning, it is often desirable to
identify and abstain on examples where the model's predictions are likely to be
incorrect. Much of the prior work on this topic focused on out-of-distribution
detection or performance metrics such as top-k accuracy. Comparatively little
attention was given to metrics such as area-under-the-curve or Cohen's Kappa,
which are extremely relevant for imbalanced datasets. Abstention strategies
aimed at top-k accuracy can produce poor results on these metrics when applied
to imbalanced datasets, even when all examples are in-distribution. We propose
a framework to address this gap. Our framework leverages the insight that
calibrated probability estimates can be used as a proxy for the true class
labels, thereby allowing us to estimate the change in an arbitrary metric if an
example were abstained on. Using this framework, we derive computationally
efficient metric-specific abstention algorithms for optimizing the sensitivity
at a target specificity level, the area under the ROC, and the weighted Cohen's
Kappa. Because our method relies only on calibrated probability estimates, we
further show that by leveraging recent work on domain adaptation under label
shift, we can generalize to test-set distributions that may have a different
class imbalance compared to the training set distribution. On various
experiments involving medical imaging, natural language processing, computer
vision and genomics, we demonstrate the effectiveness of our approach. Source
code available at https://github.com/blindauth/abstention. Colab notebooks
reproducing results available at
https://github.com/blindauth/abstention_experiments.
|
High-Dimensional Bayesian Optimization via Additive Models with
Overlapping Groups | Bayesian optimization (BO) is a popular technique for sequential black-box
function optimization, with applications including parameter tuning, robotics,
environmental monitoring, and more. One of the most important challenges in BO
is the development of algorithms that scale to high dimensions, which remains a
key open problem despite recent progress. In this paper, we consider the
approach of Kandasamy et al. (2015), in which the high-dimensional function
decomposes as a sum of lower-dimensional functions on subsets of the underlying
variables. In particular, we significantly generalize this approach by lifting
the assumption that the subsets are disjoint, and consider additive models with
arbitrary overlap among the subsets. By representing the dependencies via a
graph, we deduce an efficient message passing algorithm for optimizing the
acquisition function. In addition, we provide an algorithm for learning the
graph from samples based on Gibbs sampling. We empirically demonstrate the
effectiveness of our methods on both synthetic and real-world data.
|
The Description Length of Deep Learning Models | Solomonoff's general theory of inference and the Minimum Description Length
principle formalize Occam's razor, and hold that a good model of data is a
model that is good at losslessly compressing the data, including the cost of
describing the model itself. Deep neural networks might seem to go against this
principle given the large number of parameters to be encoded.
We demonstrate experimentally the ability of deep neural networks to compress
the training data even when accounting for parameter encoding. The compression
viewpoint originally motivated the use of variational methods in neural
networks. Unexpectedly, we found that these variational methods provide
surprisingly poor compression bounds, despite being explicitly built to
minimize such bounds. This might explain the relatively poor practical
performance of variational methods in deep learning. On the other hand, simple
incremental encoding methods yield excellent compression values on deep
networks, vindicating Solomonoff's approach.
|
On Learning Causal Structures from Non-Experimental Data without Any
Faithfulness Assumption | Consider the problem of learning, from non-experimental data, the causal
(Markov equivalence) structure of the true, unknown causal Bayesian network
(CBN) on a given, fixed set of (categorical) variables. This learning problem
is known to be so hard that there is no learning algorithm that converges to
the truth for all possible CBNs (on the given set of variables). So the
convergence property has to be sacrificed for some CBNs---but for which? In
response, the standard practice has been to design and employ learning
algorithms that secure the convergence property for at least all the CBNs that
satisfy the famous faithfulness condition, which implies sacrificing the
convergence property for some CBNs that violate the faithfulness condition
(Spirtes et al. 2000). This standard design practice can be justified by
assuming---that is, accepting on faith---that the true, unknown CBN satisfies
the faithfulness condition. But the real question is this: Is it possible to
explain, without assuming the faithfulness condition or any of its weaker
variants, why it is mandatory rather than optional to follow the standard
design practice? This paper aims to answer the above question in the
affirmative. We first define an array of modes of convergence to the truth as
desiderata that might or might not be achieved by a causal learning algorithm.
Those modes of convergence concern (i) how pervasive the domain of convergence
is on the space of all possible CBNs and (ii) how uniformly the convergence
happens. Then we prove a result to the following effect: for any learning
algorithm that tackles the causal learning problem in question, if it achieves
the best achievable mode of convergence (considered in this paper), then it
must follow the standard design practice of converging to the truth for at
least all CBNs that satisfy the faithfulness condition---it is a requirement,
not an option.
|
Robust Maximization of Non-Submodular Objectives | We study the problem of maximizing a monotone set function subject to a
cardinality constraint $k$ in the setting where some number of elements $\tau$
is deleted from the returned set. The focus of this work is on the worst-case
adversarial setting. While there exist constant-factor guarantees when the
function is submodular, there are no guarantees for non-submodular objectives.
In this work, we present a new algorithm Oblivious-Greedy and prove the first
constant-factor approximation guarantees for a wider class of non-submodular
objectives. The obtained theoretical bounds are the first constant-factor
bounds that also hold in the linear regime, i.e. when the number of deletions
$\tau$ is linear in $k$. Our bounds depend on established parameters such as
the submodularity ratio and some novel ones such as the inverse curvature. We
bound these parameters for two important objectives including support selection
and variance reduction. Finally, we numerically demonstrate the robust
performance of Oblivious-Greedy for these two objectives on various datasets.
|
i-RevNet: Deep Invertible Networks | It is widely believed that the success of deep convolutional networks is
based on progressively discarding uninformative variability about the input
with respect to the problem at hand. This is supported empirically by the
difficulty of recovering images from their hidden representations, in most
commonly used network architectures. In this paper we show via a one-to-one
mapping that this loss of information is not a necessary condition to learn
representations that generalize well on complicated problems, such as ImageNet.
Via a cascade of homeomorphic layers, we build the i-RevNet, a network that can
be fully inverted up to the final projection onto the classes, i.e. no
information is discarded. Building an invertible architecture is difficult, for
one, because the local inversion is ill-conditioned, we overcome this by
providing an explicit inverse. An analysis of i-RevNets learned representations
suggests an alternative explanation for the success of deep networks by a
progressive contraction and linear separation with depth. To shed light on the
nature of the model learned by the i-RevNet we reconstruct linear
interpolations between natural image representations.
|
Attentive Tensor Product Learning | This paper proposes a new architecture - Attentive Tensor Product Learning
(ATPL) - to represent grammatical structures in deep learning models. ATPL is a
new architecture to bridge this gap by exploiting Tensor Product
Representations (TPR), a structured neural-symbolic model developed in
cognitive science, aiming to integrate deep learning with explicit language
structures and rules. The key ideas of ATPL are: 1) unsupervised learning of
role-unbinding vectors of words via TPR-based deep neural network; 2) employing
attention modules to compute TPR; and 3) integration of TPR with typical deep
learning architectures including Long Short-Term Memory (LSTM) and Feedforward
Neural Network (FFNN). The novelty of our approach lies in its ability to
extract the grammatical structure of a sentence by using role-unbinding
vectors, which are obtained in an unsupervised manner. This ATPL approach is
applied to 1) image captioning, 2) part of speech (POS) tagging, and 3)
constituency parsing of a sentence. Experimental results demonstrate the
effectiveness of the proposed approach.
|
An Efficient Semismooth Newton Based Algorithm for Convex Clustering | Clustering may be the most fundamental problem in unsupervised learning which
is still active in machine learning research because its importance in many
applications. Popular methods like K-means, may suffer from instability as they
are prone to get stuck in its local minima. Recently, the sum-of-norms (SON)
model (also known as clustering path), which is a convex relaxation of
hierarchical clustering model, has been proposed in [7] and [5] Although
numerical algorithms like ADMM and AMA are proposed to solve convex clustering
model [2], it is known to be very challenging to solve large-scale problems. In
this paper, we propose a semi-smooth Newton based augmented Lagrangian method
for large-scale convex clustering problems. Extensive numerical experiments on
both simulated and real data demonstrate that our algorithm is highly efficient
and robust for solving large-scale problems. Moreover, the numerical results
also show the superior performance and scalability of our algorithm compared to
existing first-order methods.
|
Do Less, Get More: Streaming Submodular Maximization with Subsampling | In this paper, we develop the first one-pass streaming algorithm for
submodular maximization that does not evaluate the entire stream even once. By
carefully subsampling each element of data stream, our algorithm enjoys the
tightest approximation guarantees in various settings while having the smallest
memory footprint and requiring the lowest number of function evaluations. More
specifically, for a monotone submodular function and a $p$-matchoid constraint,
our randomized algorithm achieves a $4p$ approximation ratio (in expectation)
with $O(k)$ memory and $O(km/p)$ queries per element ($k$ is the size of the
largest feasible solution and $m$ is the number of matroids used to define the
constraint). For the non-monotone case, our approximation ratio increases only
slightly to $4p+2-o(1)$. To the best or our knowledge, our algorithm is the
first that combines the benefits of streaming and subsampling in a novel way in
order to truly scale submodular maximization to massive machine learning
problems. To showcase its practicality, we empirically evaluated the
performance of our algorithm on a video summarization application and observed
that it outperforms the state-of-the-art algorithm by up to fifty fold, while
maintaining practically the same utility.
|
Learning of Optimal Forecast Aggregation in Partial Evidence
Environments | We consider the forecast aggregation problem in repeated settings, where the
forecasts are done on a binary event. At each period multiple experts provide
forecasts about an event. The goal of the aggregator is to aggregate those
forecasts into a subjective accurate forecast. We assume that experts are
Bayesian; namely they share a common prior, each expert is exposed to some
evidence, and each expert applies Bayes rule to deduce his forecast. The
aggregator is ignorant with respect to the information structure (i.e.,
distribution over evidence) according to which experts make their prediction.
The aggregator observes the experts' forecasts only. At the end of each period
the actual state is realized. We focus on the question whether the aggregator
can learn to aggregate optimally the forecasts of the experts, where the
optimal aggregation is the Bayesian aggregation that takes into account all the
information (evidence) in the system.
We consider the class of partial evidence information structures, where each
expert is exposed to a different subset of conditionally independent signals.
Our main results are positive; We show that optimal aggregation can be learned
in polynomial time in a quite wide range of instances of the partial evidence
environments. We provide a tight characterization of the instances where
learning is possible and impossible.
|
Out-distribution training confers robustness to deep neural networks | The easiness at which adversarial instances can be generated in deep neural
networks raises some fundamental questions on their functioning and concerns on
their use in critical systems. In this paper, we draw a connection between
over-generalization and adversaries: a possible cause of adversaries lies in
models designed to make decisions all over the input space, leading to
inappropriate high-confidence decisions in parts of the input space not
represented in the training set. We empirically show an augmented neural
network, which is not trained on any types of adversaries, can increase the
robustness by detecting black-box one-step adversaries, i.e. assimilated to
out-distribution samples, and making generation of white-box one-step
adversaries harder.
|
On Estimating Multi-Attribute Choice Preferences using Private Signals
and Matrix Factorization | Revealed preference theory studies the possibility of modeling an agent's
revealed preferences and the construction of a consistent utility function.
However, modeling agent's choices over preference orderings is not always
practical and demands strong assumptions on human rationality and
data-acquisition abilities. Therefore, we propose a simple generative choice
model where agents are assumed to generate the choice probabilities based on
latent factor matrices that capture their choice evaluation across multiple
attributes. Since the multi-attribute evaluation is typically hidden within the
agent's psyche, we consider a signaling mechanism where agents are provided
with choice information through private signals, so that the agent's choices
provide more insight about his/her latent evaluation across multiple
attributes. We estimate the choice model via a novel multi-stage matrix
factorization algorithm that minimizes the average deviation of the factor
estimates from choice data. Simulation results are presented to validate the
estimation performance of our proposed algorithm.
|
Local Differential Privacy for Evolving Data | There are now several large scale deployments of differential privacy used to
collect statistical information about users. However, these deployments
periodically recollect the data and recompute the statistics using algorithms
designed for a single use. As a result, these systems do not provide meaningful
privacy guarantees over long time scales. Moreover, existing techniques to
mitigate this effect do not apply in the "local model" of differential privacy
that these systems use.
In this paper, we introduce a new technique for local differential privacy
that makes it possible to maintain up-to-date statistics over time, with
privacy guarantees that degrade only in the number of changes in the underlying
distribution rather than the number of collection periods. We use our technique
for tracking a changing statistic in the setting where users are partitioned
into an unknown collection of groups, and at every time period each user draws
a single bit from a common (but changing) group-specific distribution. We also
provide an application to frequency and heavy-hitter estimation.
|
Deep BCD-Net Using Identical Encoding-Decoding CNN Structures for
Iterative Image Recovery | In "extreme" computational imaging that collects extremely undersampled or
noisy measurements, obtaining an accurate image within a reasonable computing
time is challenging. Incorporating image mapping convolutional neural networks
(CNN) into iterative image recovery has great potential to resolve this issue.
This paper 1) incorporates image mapping CNN using identical convolutional
kernels in both encoders and decoders into a block coordinate descent (BCD)
signal recovery method and 2) applies alternating direction method of
multipliers to train the aforementioned image mapping CNN. We refer to the
proposed recurrent network as BCD-Net using identical encoding-decoding CNN
structures. Numerical experiments show that, for a) denoising low
signal-to-noise-ratio images and b) extremely undersampled magnetic resonance
imaging, the proposed BCD-Net achieves significantly more accurate image
recovery, compared to BCD-Net using distinct encoding-decoding structures
and/or the conventional image recovery model using both wavelets and total
variation.
|
Adaptive Sampling for Coarse Ranking | We consider the problem of active coarse ranking, where the goal is to sort
items according to their means into clusters of pre-specified sizes, by
adaptively sampling from their reward distributions. This setting is useful in
many social science applications involving human raters and the approximate
rank of every item is desired. Approximate or coarse ranking can significantly
reduce the number of ratings required in comparison to the number needed to
find an exact ranking. We propose a computationally efficient PAC algorithm
LUCBRank for coarse ranking, and derive an upper bound on its sample
complexity. We also derive a nearly matching distribution-dependent lower
bound. Experiments on synthetic as well as real-world data show that LUCBRank
performs better than state-of-the-art baseline methods, even when these methods
have the advantage of knowing the underlying parametric model.
|
Neural Architecture Search with Bayesian Optimisation and Optimal
Transport | Bayesian Optimisation (BO) refers to a class of methods for global
optimisation of a function $f$ which is only accessible via point evaluations.
It is typically used in settings where $f$ is expensive to evaluate. A common
use case for BO in machine learning is model selection, where it is not
possible to analytically model the generalisation performance of a statistical
model, and we resort to noisy and expensive training and validation procedures
to choose the best model. Conventional BO methods have focused on Euclidean and
categorical domains, which, in the context of model selection, only permits
tuning scalar hyper-parameters of machine learning algorithms. However, with
the surge of interest in deep learning, there is an increasing demand to tune
neural network \emph{architectures}. In this work, we develop NASBOT, a
Gaussian process based BO framework for neural architecture search. To
accomplish this, we develop a distance metric in the space of neural network
architectures which can be computed efficiently via an optimal transport
program. This distance might be of independent interest to the deep learning
community as it may find applications outside of BO. We demonstrate that NASBOT
outperforms other alternatives for architecture search in several cross
validation based model selection tasks on multi-layer perceptrons and
convolutional neural networks.
|
AutoPrognosis: Automated Clinical Prognostic Modeling via Bayesian
Optimization with Structured Kernel Learning | Clinical prognostic models derived from largescale healthcare data can inform
critical diagnostic and therapeutic decisions. To enable off-theshelf usage of
machine learning (ML) in prognostic research, we developed AUTOPROGNOSIS: a
system for automating the design of predictive modeling pipelines tailored for
clinical prognosis. AUTOPROGNOSIS optimizes ensembles of pipeline
configurations efficiently using a novel batched Bayesian optimization (BO)
algorithm that learns a low-dimensional decomposition of the pipelines
high-dimensional hyperparameter space in concurrence with the BO procedure.
This is achieved by modeling the pipelines performances as a black-box function
with a Gaussian process prior, and modeling the similarities between the
pipelines baseline algorithms via a sparse additive kernel with a Dirichlet
prior. Meta-learning is used to warmstart BO with external data from similar
patient cohorts by calibrating the priors using an algorithm that mimics the
empirical Bayes method. The system automatically explains its predictions by
presenting the clinicians with logical association rules that link patients
features to predicted risk strata. We demonstrate the utility of AUTOPROGNOSIS
using 10 major patient cohorts representing various aspects of cardiovascular
patient care.
|
Actively Avoiding Nonsense in Generative Models | A generative model may generate utter nonsense when it is fit to maximize the
likelihood of observed data. This happens due to "model error," i.e., when the
true data generating distribution does not fit within the class of generative
models being learned. To address this, we propose a model of active
distribution learning using a binary invalidity oracle that identifies some
examples as clearly invalid, together with random positive examples sampled
from the true distribution. The goal is to maximize the likelihood of the
positive examples subject to the constraint of (almost) never generating
examples labeled invalid by the oracle. Guarantees are agnostic compared to a
class of probability distributions. We show that, while proper learning often
requires exponentially many queries to the invalidity oracle, improper
distribution learning can be done using polynomially many queries.
|
Continual Reinforcement Learning with Complex Synapses | Unlike humans, who are capable of continual learning over their lifetimes,
artificial neural networks have long been known to suffer from a phenomenon
known as catastrophic forgetting, whereby new learning can lead to abrupt
erasure of previously acquired knowledge. Whereas in a neural network the
parameters are typically modelled as scalar values, an individual synapse in
the brain comprises a complex network of interacting biochemical components
that evolve at different timescales. In this paper, we show that by equipping
tabular and deep reinforcement learning agents with a synaptic model that
incorporates this biological complexity (Benna & Fusi, 2016), catastrophic
forgetting can be mitigated at multiple timescales. In particular, we find that
as well as enabling continual learning across sequential training of two simple
tasks, it can also be used to overcome within-task forgetting by reducing the
need for an experience replay database.
|
Steering Social Activity: A Stochastic Optimal Control Point Of View | User engagement in online social networking depends critically on the level
of social activity in the corresponding platform--the number of online actions,
such as posts, shares or replies, taken by their users. Can we design
data-driven algorithms to increase social activity? At a user level, such
algorithms may increase activity by helping users decide when to take an action
to be more likely to be noticed by their peers. At a network level, they may
increase activity by incentivizing a few influential users to take more
actions, which in turn will trigger additional actions by other users. In this
paper, we model social activity using the framework of marked temporal point
processes, derive an alternate representation of these processes using
stochastic differential equations (SDEs) with jumps and, exploiting this
alternate representation, develop two efficient online algorithms with provable
guarantees to steer social activity both at a user and at a network level. In
doing so, we establish a previously unexplored connection between optimal
control of jump SDEs and doubly stochastic marked temporal point processes,
which is of independent interest. Finally, we experiment both with synthetic
and real data gathered from Twitter and show that our algorithms consistently
steer social activity more effectively than the state of the art.
|
Meta-Reinforcement Learning of Structured Exploration Strategies | Exploration is a fundamental challenge in reinforcement learning (RL). Many
of the current exploration methods for deep RL use task-agnostic objectives,
such as information gain or bonuses based on state visitation. However, many
practical applications of RL involve learning more than a single task, and
prior tasks can be used to inform how exploration should be performed in new
tasks. In this work, we explore how prior tasks can inform an agent about how
to explore effectively in new situations. We introduce a novel gradient-based
fast adaptation algorithm -- model agnostic exploration with structured noise
(MAESN) -- to learn exploration strategies from prior experience. The prior
experience is used both to initialize a policy and to acquire a latent
exploration space that can inject structured stochasticity into a policy,
producing exploration strategies that are informed by prior knowledge and are
more effective than random action-space noise. We show that MAESN is more
effective at learning exploration strategies when compared to prior meta-RL
methods, RL without learned exploration strategies, and task-agnostic
exploration methods. We evaluate our method on a variety of simulated tasks:
locomotion with a wheeled robot, locomotion with a quadrupedal walker, and
object manipulation.
|
Attack Strength vs. Detectability Dilemma in Adversarial Machine
Learning | As the prevalence and everyday use of machine learning algorithms, along with
our reliance on these algorithms grow dramatically, so do the efforts to attack
and undermine these algorithms with malicious intent, resulting in a growing
interest in adversarial machine learning. A number of approaches have been
developed that can render a machine learning algorithm ineffective through
poisoning or other types of attacks. Most attack algorithms typically use
sophisticated optimization approaches, whose objective function is designed to
cause maximum damage with respect to accuracy and performance of the algorithm
with respect to some task. In this effort, we show that while such an objective
function is indeed brutally effective in causing maximum damage on an embedded
feature selection task, it often results in an attack mechanism that can be
easily detected with an embarrassingly simple novelty or outlier detection
algorithm. We then propose an equally simple yet elegant solution by adding a
regularization term to the attacker's objective function that penalizes
outlying attack points.
|
On the Connection Between Learning Two-Layers Neural Networks and Tensor
Decomposition | We establish connections between the problem of learning a two-layer neural
network and tensor decomposition. We consider a model with feature vectors
$\boldsymbol x \in \mathbb R^d$, $r$ hidden units with weights $\{\boldsymbol
w_i\}_{1\le i \le r}$ and output $y\in \mathbb R$, i.e., $y=\sum_{i=1}^r
\sigma( \boldsymbol w_i^{\mathsf T}\boldsymbol x)$, with activation functions
given by low-degree polynomials. In particular, if $\sigma(x) =
a_0+a_1x+a_3x^3$, we prove that no polynomial-time learning algorithm can
outperform the trivial predictor that assigns to each example the response
variable $\mathbb E(y)$, when $d^{3/2}\ll r\ll d^2$. Our conclusion holds for a
`natural data distribution', namely standard Gaussian feature vectors
$\boldsymbol x$, and output distributed according to a two-layer neural network
with random isotropic weights, and under a certain complexity-theoretic
assumption on tensor decomposition. Roughly speaking, we assume that no
polynomial-time algorithm can substantially outperform current methods for
tensor decomposition based on the sum-of-squares hierarchy.
We also prove generalizations of this statement for higher degree polynomial
activations, and non-random weight vectors. Remarkably, several existing
algorithms for learning two-layer networks with rigorous guarantees are based
on tensor decomposition. Our results support the idea that this is indeed the
core computational difficulty in learning such networks, under the stated
generative model for the data. As a side result, we show that under this model
learning the network requires accurate learning of its weights, a property that
does not hold in a more general setting.
|
Bayesian Incremental Learning for Deep Neural Networks | In industrial machine learning pipelines, data often arrive in parts.
Particularly in the case of deep neural networks, it may be too expensive to
train the model from scratch each time, so one would rather use a previously
learned model and the new data to improve performance. However, deep neural
networks are prone to getting stuck in a suboptimal solution when trained on
only new data as compared to the full dataset. Our work focuses on a continuous
learning setup where the task is always the same and new parts of data arrive
sequentially. We apply a Bayesian approach to update the posterior
approximation with each new piece of data and find this method to outperform
the traditional approach in our experiments.
|
On the Statistical Challenges of Echo State Networks and Some Potential
Remedies | Echo state networks are powerful recurrent neural networks. However, they are
often unstable and shaky, making the process of finding an good ESN for a
specific dataset quite hard. Obtaining a superb accuracy by using the Echo
State Network is a challenging task. We create, develop and implement a family
of predictably optimal robust and stable ensemble of Echo State Networks via
regularizing the training and perturbing the input. Furthermore, several
distributions of weights have been tried based on the shape to see if the shape
of the distribution has the impact for reducing the error. We found ESN can
track in short term for most dataset, but it collapses in the long run.
Short-term tracking with large size reservoir enables ESN to perform strikingly
with superior prediction. Based on this scenario, we go a further step to
aggregate many of ESNs into an ensemble to lower the variance and stabilize the
system by stochastic replications and bootstrapping of input data.
|
Stochastic Variance-Reduced Cubic Regularization for Nonconvex
Optimization | Cubic regularization (CR) is an optimization method with emerging popularity
due to its capability to escape saddle points and converge to second-order
stationary solutions for nonconvex optimization. However, CR encounters a high
sample complexity issue for finite-sum problems with a large data size.
%Various inexact variants of CR have been proposed to improve the sample
complexity. In this paper, we propose a stochastic variance-reduced
cubic-regularization (SVRC) method under random sampling, and study its
convergence guarantee as well as sample complexity. We show that the iteration
complexity of SVRC for achieving a second-order stationary solution within
$\epsilon$ accuracy is $O(\epsilon^{-3/2})$, which matches the state-of-art
result on CR types of methods. Moreover, our proposed variance reduction scheme
significantly reduces the per-iteration sample complexity. The resulting total
Hessian sample complexity of our SVRC is ${\Oc}(N^{2/3} \epsilon^{-3/2})$,
which outperforms the state-of-art result by a factor of $O(N^{2/15})$. We also
study our SVRC under random sampling without replacement scheme, which yields a
lower per-iteration sample complexity, and hence justifies its practical
applicability.
|
Scalable Label Propagation for Multi-relational Learning on the Tensor
Product of Graphs | Multi-relational learning on knowledge graphs infers high-order relations
among the entities across the graphs. This learning task can be solved by label
propagation on the tensor product of the knowledge graphs to learn the
high-order relations as a tensor. In this paper, we generalize a widely used
label propagation model to the normalized tensor product graph, and propose an
optimization formulation and a scalable Low-rank Tensor-based Label Propagation
algorithm (LowrankTLP) to infer multi-relations for two learning tasks,
hyperlink prediction and multiple graph alignment. The optimization formulation
minimizes the upper bound of the noisy tensor estimation error for multiple
graph alignment, by learning with a subset of the eigen-pairs in the spectrum
of the normalized tensor product graph. We also provide a data-dependent
transductive Rademacher bound for binary hyperlink prediction. We accelerate
LowrankTLP with parallel tensor computation which enables label propagation on
a tensor product of 100 graphs each of size 1000 in less than half hour in the
simulation. LowrankTLP was also applied to predicting the author-paper-venue
hyperlinks in publication records, alignment of segmented regions across up to
26 CT-scan images and alignment of protein-protein interaction networks across
multiple species. The experiments demonstrate that LowrankTLP indeed well
approximates the original label propagation with better scalability and
accuracy.
|
Generic Coreset for Scalable Learning of Monotonic Kernels: Logistic
Regression, Sigmoid and more | Coreset (or core-set) is a small weighted \emph{subset} $Q$ of an input set
$P$ with respect to a given \emph{monotonic} function
$f:\mathbb{R}\to\mathbb{R}$ that \emph{provably} approximates its fitting loss
$\sum_{p\in P}f(p\cdot x)$ to \emph{any} given $x\in\mathbb{R}^d$. Using $Q$ we
can obtain approximation of $x^*$ that minimizes this loss, by running
\emph{existing} optimization algorithms on $Q$. In this work we provide: (i) A
lower bound which proves that there are sets with no coresets smaller than
$n=|P|$ for general monotonic loss functions. (ii) A proof that, under a
natural assumption that holds e.g. for logistic regression and the sigmoid
activation functions, a small coreset exists for \emph{any} input $P$. (iii) A
generic coreset construction algorithm that computes such a small coreset $Q$
in $O(nd+n\log n)$ time, and (iv) Experimental results which demonstrate that
our coresets are effective and are much smaller in practice than predicted in
theory.
|
Interpreting Neural Network Judgments via Minimal, Stable, and Symbolic
Corrections | We present a new algorithm to generate minimal, stable, and symbolic
corrections to an input that will cause a neural network with ReLU activations
to change its output. We argue that such a correction is a useful way to
provide feedback to a user when the network's output is different from a
desired output. Our algorithm generates such a correction by solving a series
of linear constraint satisfaction problems. The technique is evaluated on three
neural network models: one predicting whether an applicant will pay a mortgage,
one predicting whether a first-order theorem can be proved efficiently by a
solver using certain heuristics, and the final one judging whether a drawing is
an accurate rendition of a canonical drawing of a cat.
|
3LC: Lightweight and Effective Traffic Compression for Distributed
Machine Learning | The performance and efficiency of distributed machine learning (ML) depends
significantly on how long it takes for nodes to exchange state changes.
Overly-aggressive attempts to reduce communication often sacrifice final model
accuracy and necessitate additional ML techniques to compensate for this loss,
limiting their generality. Some attempts to reduce communication incur high
computation overhead, which makes their performance benefits visible only over
slow networks.
We present 3LC, a lossy compression scheme for state change traffic that
strikes balance between multiple goals: traffic reduction, accuracy,
computation overhead, and generality. It combines three new
techniques---3-value quantization with sparsity multiplication, quartic
encoding, and zero-run encoding---to leverage strengths of quantization and
sparsification techniques and avoid their drawbacks. It achieves a data
compression ratio of up to 39--107X, almost the same test accuracy of trained
models, and high compression speed. Distributed ML frameworks can employ 3LC
without modifications to existing ML algorithms. Our experiments show that 3LC
reduces wall-clock training time of ResNet-110--based image classifiers for
CIFAR-10 on a 10-GPU cluster by up to 16--23X compared to TensorFlow's baseline
design.
|
Direct Learning to Rank and Rerank | Learning-to-rank techniques have proven to be extremely useful for
prioritization problems, where we rank items in order of their estimated
probabilities, and dedicate our limited resources to the top-ranked items. This
work exposes a serious problem with the state of learning-to-rank algorithms,
which is that they are based on convex proxies that lead to poor
approximations. We then discuss the possibility of "exact" reranking algorithms
based on mathematical programming. We prove that a relaxed version of the
"exact" problem has the same optimal solution, and provide an empirical
analysis.
|
A Study into the similarity in generator and discriminator in GAN
architecture | One popular generative model that has high-quality results is the Generative
Adversarial Networks(GAN). This type of architecture consists of two separate
networks that play against each other. The generator creates an output from the
input noise that is given to it. The discriminator has the task of determining
if the input to it is real or fake. This takes place constantly eventually
leads to the generator modeling the target distribution. This paper includes a
study into the actual weights learned by the network and a study into the
similarity of the discriminator and generator networks. The paper also tries to
leverage the similarity between these networks and shows that indeed both the
networks may have a similar structure with experimental evidence with a novel
shared architecture.
|
Breaking the gridlock in Mixture-of-Experts: Consistent and Efficient
Algorithms | Mixture-of-Experts (MoE) is a widely popular model for ensemble learning and
is a basic building block of highly successful modern neural networks as well
as a component in Gated Recurrent Units (GRU) and Attention networks. However,
present algorithms for learning MoE including the EM algorithm, and gradient
descent are known to get stuck in local optima. From a theoretical viewpoint,
finding an efficient and provably consistent algorithm to learn the parameters
remains a long standing open problem for more than two decades. In this paper,
we introduce the first algorithm that learns the true parameters of a MoE model
for a wide class of non-linearities with global consistency guarantees. While
existing algorithms jointly or iteratively estimate the expert parameters and
the gating paramters in the MoE, we propose a novel algorithm that breaks the
deadlock and can directly estimate the expert parameters by sensing its echo in
a carefully designed cross-moment tensor between the inputs and the output.
Once the experts are known, the recovery of gating parameters still requires an
EM algorithm; however, we show that the EM algorithm for this simplified
problem, unlike the joint EM algorithm, converges to the true parameters. We
empirically validate our algorithm on both the synthetic and real data sets in
a variety of settings, and show superior performance to standard baselines.
|
Generalization in Machine Learning via Analytical Learning Theory | This paper introduces a novel measure-theoretic theory for machine learning
that does not require statistical assumptions. Based on this theory, a new
regularization method in deep learning is derived and shown to outperform
previous methods in CIFAR-10, CIFAR-100, and SVHN. Moreover, the proposed
theory provides a theoretical basis for a family of practically successful
regularization methods in deep learning. We discuss several consequences of our
results on one-shot learning, representation learning, deep learning, and
curriculum learning. Unlike statistical learning theory, the proposed learning
theory analyzes each problem instance individually via measure theory, rather
than a set of problem instances via statistics. As a result, it provides
different types of results and insights when compared to statistical learning
theory.
|
Active Learning with Partial Feedback | While many active learning papers assume that the learner can simply ask for
a label and receive it, real annotation often presents a mismatch between the
form of a label (say, one among many classes), and the form of an annotation
(typically yes/no binary feedback). To annotate examples corpora for multiclass
classification, we might need to ask multiple yes/no questions, exploiting a
label hierarchy if one is available. To address this more realistic setting, we
propose active learning with partial feedback (ALPF), where the learner must
actively choose both which example to label and which binary question to ask.
At each step, the learner selects an example, asking if it belongs to a chosen
(possibly composite) class. Each answer eliminates some classes, leaving the
learner with a partial label. The learner may then either ask more questions
about the same example (until an exact label is uncovered) or move on
immediately, leaving the first example partially labeled. Active learning with
partial labels requires (i) a sampling strategy to choose (example, class)
pairs, and (ii) learning from partial labels between rounds. Experiments on
Tiny ImageNet demonstrate that our most effective method improves 26%
(relative) in top-1 classification accuracy compared to i.i.d. baselines and
standard active learners given 30% of the annotation budget that would be
required (naively) to annotate the dataset. Moreover, ALPF-learners fully
annotate TinyImageNet at 42% lower cost. Surprisingly, we observe that
accounting for per-example annotation costs can alter the conventional wisdom
that active learners should solicit labels for hard examples.
|
Learning to Play with Intrinsically-Motivated Self-Aware Agents | Infants are experts at playing, with an amazing ability to generate novel
structured behaviors in unstructured environments that lack clear extrinsic
reward signals. We seek to mathematically formalize these abilities using a
neural network that implements curiosity-driven intrinsic motivation. Using a
simple but ecologically naturalistic simulated environment in which an agent
can move and interact with objects it sees, we propose a "world-model" network
that learns to predict the dynamic consequences of the agent's actions.
Simultaneously, we train a separate explicit "self-model" that allows the agent
to track the error map of its own world-model, and then uses the self-model to
adversarially challenge the developing world-model. We demonstrate that this
policy causes the agent to explore novel and informative interactions with its
environment, leading to the generation of a spectrum of complex behaviors,
including ego-motion prediction, object attention, and object gathering.
Moreover, the world-model that the agent learns supports improved performance
on object dynamics prediction, detection, localization and recognition tasks.
Taken together, our results are initial steps toward creating flexible
autonomous agents that self-supervise in complex novel physical environments.
|
Scaling-up Split-Merge MCMC with Locality Sensitive Sampling (LSS) | Split-Merge MCMC (Monte Carlo Markov Chain) is one of the essential and
popular variants of MCMC for problems when an MCMC state consists of an unknown
number of components. It is well known that state-of-the-art methods for
split-merge MCMC do not scale well. Strategies for rapid mixing requires smart
and informative proposals to reduce the rejection rate. However, all known
smart proposals involve expensive operations to suggest informative
transitions. As a result, the cost of each iteration is prohibitive for massive
scale datasets. It is further known that uninformative but computationally
efficient proposals, such as random split-merge, leads to extremely slow
convergence. This tradeoff between mixing time and per update cost seems hard
to get around.
In this paper, we show a sweet spot. We leverage some unique properties of
weighted MinHash, which is a popular LSH, to design a novel class of
split-merge proposals which are significantly more informative than random
sampling but at the same time efficient to compute. Overall, we obtain a
superior tradeoff between convergence and per update cost. As a direct
consequence, our proposals are around 6X faster than the state-of-the-art
sampling methods on two large real datasets KDDCUP and PubMed with several
millions of entities and thousands of clusters.
|
Emergence of Structured Behaviors from Curiosity-Based Intrinsic
Motivation | Infants are experts at playing, with an amazing ability to generate novel
structured behaviors in unstructured environments that lack clear extrinsic
reward signals. We seek to replicate some of these abilities with a neural
network that implements curiosity-driven intrinsic motivation. Using a simple
but ecologically naturalistic simulated environment in which the agent can move
and interact with objects it sees, the agent learns a world model predicting
the dynamic consequences of its actions. Simultaneously, the agent learns to
take actions that adversarially challenge the developing world model, pushing
the agent to explore novel and informative interactions with its environment.
We demonstrate that this policy leads to the self-supervised emergence of a
spectrum of complex behaviors, including ego motion prediction, object
attention, and object gathering. Moreover, the world model that the agent
learns supports improved performance on object dynamics prediction and
localization tasks. Our results are a proof-of-principle that computational
models of intrinsic motivation might account for key features of developmental
visuomotor learning in infants.
|
Data-Driven Forecasting of High-Dimensional Chaotic Systems with Long
Short-Term Memory Networks | We introduce a data-driven forecasting method for high-dimensional chaotic
systems using long short-term memory (LSTM) recurrent neural networks. The
proposed LSTM neural networks perform inference of high-dimensional dynamical
systems in their reduced order space and are shown to be an effective set of
nonlinear approximators of their attractor. We demonstrate the forecasting
performance of the LSTM and compare it with Gaussian processes (GPs) in time
series obtained from the Lorenz 96 system, the Kuramoto-Sivashinsky equation
and a prototype climate model. The LSTM networks outperform the GPs in
short-term forecasting accuracy in all applications considered. A hybrid
architecture, extending the LSTM with a mean stochastic model (MSM-LSTM), is
proposed to ensure convergence to the invariant measure. This novel hybrid
method is fully data-driven and extends the forecasting capabilities of LSTM
networks.
|
Spectrally approximating large graphs with smaller graphs | How does coarsening affect the spectrum of a general graph? We provide
conditions such that the principal eigenvalues and eigenspaces of a coarsened
and original graph Laplacian matrices are close. The achieved approximation is
shown to depend on standard graph-theoretic properties, such as the degree and
eigenvalue distributions, as well as on the ratio between the coarsened and
actual graph sizes. Our results carry implications for learning methods that
utilize coarsening. For the particular case of spectral clustering, they imply
that coarse eigenvectors can be used to derive good quality assignments even
without refinement---this phenomenon was previously observed, but lacked formal
justification.
|
Adversarial classification: An adversarial risk analysis approach | Classification problems in security settings are usually contemplated as
confrontations in which one or more adversaries try to fool a classifier to
obtain a benefit. Most approaches to such adversarial classification problems
have focused on game theoretical ideas with strong underlying common knowledge
assumptions, which are actually not realistic in security domains. We provide
an alternative framework to such problem based on adversarial risk analysis,
which we illustrate with several examples. Computational and implementation
issues are discussed.
|
Learning Integral Representations of Gaussian Processes | We propose a representation of Gaussian processes (GPs) based on powers of
the integral operator defined by a kernel function, we call these stochastic
processes integral Gaussian processes (IGPs). Sample paths from IGPs are
functions contained within the reproducing kernel Hilbert space (RKHS) defined
by the kernel function, in contrast sample paths from the standard GP are not
functions within the RKHS. We develop computationally efficient non-parametric
regression models based on IGPs. The main innovation in our regression
algorithm is the construction of a low dimensional subspace that captures the
information most relevant to explaining variation in the response. We use ideas
from supervised dimension reduction to compute this subspace. The result of
using the construction we propose involves significant improvements in the
computational complexity of estimating kernel hyper-parameters as well as
reducing the prediction variance.
|
The Many Faces of Exponential Weights in Online Learning | A standard introduction to online learning might place Online Gradient
Descent at its center and then proceed to develop generalizations and
extensions like Online Mirror Descent and second-order methods. Here we explore
the alternative approach of putting Exponential Weights (EW) first. We show
that many standard methods and their regret bounds then follow as a special
case by plugging in suitable surrogate losses and playing the EW posterior
mean. For instance, we easily recover Online Gradient Descent by using EW with
a Gaussian prior on linearized losses, and, more generally, all instances of
Online Mirror Descent based on regular Bregman divergences also correspond to
EW with a prior that depends on the mirror map. Furthermore, appropriate
quadratic surrogate losses naturally give rise to Online Gradient Descent for
strongly convex losses and to Online Newton Step. We further interpret several
recent adaptive methods (iProd, Squint, and a variation of Coin Betting for
experts) as a series of closely related reductions to exp-concave surrogate
losses that are then handled by Exponential Weights. Finally, a benefit of our
EW interpretation is that it opens up the possibility of sampling from the EW
posterior distribution instead of playing the mean. As already observed by
Bubeck and Eldan, this recovers the best-known rate in Online Bandit Linear
Optimization.
|
Clipped Action Policy Gradient | Many continuous control tasks have bounded action spaces. When policy
gradient methods are applied to such tasks, out-of-bound actions need to be
clipped before execution, while policies are usually optimized as if the
actions are not clipped. We propose a policy gradient estimator that exploits
the knowledge of actions being clipped to reduce the variance in estimation. We
prove that our estimator, named clipped action policy gradient (CAPG), is
unbiased and achieves lower variance than the conventional estimator that
ignores action bounds. Experimental results demonstrate that CAPG generally
outperforms the conventional estimator, indicating that it is a better policy
gradient estimator for continuous control tasks. The source code is available
at https://github.com/pfnet-research/capg.
|
Continual Lifelong Learning with Neural Networks: A Review | Humans and animals have the ability to continually acquire, fine-tune, and
transfer knowledge and skills throughout their lifespan. This ability, referred
to as lifelong learning, is mediated by a rich set of neurocognitive mechanisms
that together contribute to the development and specialization of our
sensorimotor skills as well as to long-term memory consolidation and retrieval.
Consequently, lifelong learning capabilities are crucial for autonomous agents
interacting in the real world and processing continuous streams of information.
However, lifelong learning remains a long-standing challenge for machine
learning and neural network models since the continual acquisition of
incrementally available information from non-stationary data distributions
generally leads to catastrophic forgetting or interference. This limitation
represents a major drawback for state-of-the-art deep neural network models
that typically learn representations from stationary batches of training data,
thus without accounting for situations in which information becomes
incrementally available over time. In this review, we critically summarize the
main challenges linked to lifelong learning for artificial learning systems and
compare existing neural network approaches that alleviate, to different
extents, catastrophic forgetting. We discuss well-established and emerging
research motivated by lifelong learning factors in biological systems such as
structural plasticity, memory replay, curriculum and transfer learning,
intrinsic motivation, and multisensory integration.
|
Information Theoretic Co-Training | This paper introduces an information theoretic co-training objective for
unsupervised learning. We consider the problem of predicting the future. Rather
than predict future sensations (image pixels or sound waves) we predict
"hypotheses" to be confirmed by future sensations. More formally, we assume a
population distribution on pairs $(x,y)$ where we can think of $x$ as a past
sensation and $y$ as a future sensation. We train both a predictor model
$P_\Phi(z|x)$ and a confirmation model $P_\Psi(z|y)$ where we view $z$ as
hypotheses (when predicted) or facts (when confirmed). For a population
distribution on pairs $(x,y)$ we focus on the problem of measuring the mutual
information between $x$ and $y$. By the data processing inequality this mutual
information is at least as large as the mutual information between $x$ and $z$
under the distribution on triples $(x,z,y)$ defined by the confirmation model
$P_\Psi(z|y)$. The information theoretic training objective for $P_\Phi(z|x)$
and $P_\Psi(z|y)$ can be viewed as a form of co-training where we want the
prediction from $x$ to match the confirmation from $y$.
|
Universal Hypothesis Testing with Kernels: Asymptotically Optimal Tests
for Goodness of Fit | We characterize the asymptotic performance of nonparametric goodness of fit
testing. The exponential decay rate of the type-II error probability is used as
the asymptotic performance metric, and a test is optimal if it achieves the
maximum rate subject to a constant level constraint on the type-I error
probability. We show that two classes of Maximum Mean Discrepancy (MMD) based
tests attain this optimality on $\mathbb R^d$, while the quadratic-time Kernel
Stein Discrepancy (KSD) based tests achieve the maximum exponential decay rate
under a relaxed level constraint. Under the same performance metric, we proceed
to show that the quadratic-time MMD based two-sample tests are also optimal for
general two-sample problems, provided that kernels are bounded continuous and
characteristic. Key to our approach are Sanov's theorem from large deviation
theory and the weak metrizable properties of the MMD and KSD.
|
Smooth Loss Functions for Deep Top-k Classification | The top-k error is a common measure of performance in machine learning and
computer vision. In practice, top-k classification is typically performed with
deep neural networks trained with the cross-entropy loss. Theoretical results
indeed suggest that cross-entropy is an optimal learning objective for such a
task in the limit of infinite data. In the context of limited and noisy data
however, the use of a loss function that is specifically designed for top-k
classification can bring significant improvements. Our empirical evidence
suggests that the loss function must be smooth and have non-sparse gradients in
order to work well with deep neural networks. Consequently, we introduce a
family of smoothed loss functions that are suited to top-k optimization via
deep learning. The widely used cross-entropy is a special case of our family.
Evaluating our smooth loss functions is computationally challenging: a na\"ive
algorithm would require $\mathcal{O}(\binom{n}{k})$ operations, where n is the
number of classes. Thanks to a connection to polynomial algebra and a
divide-and-conquer approach, we provide an algorithm with a time complexity of
$\mathcal{O}(k n)$. Furthermore, we present a novel approximation to obtain
fast and stable algorithms on GPUs with single floating point precision. We
compare the performance of the cross-entropy loss and our margin-based losses
in various regimes of noise and data size, for the predominant use case of k=5.
Our investigation reveals that our loss is more robust to noise and overfitting
than cross-entropy.
|
Ordered Preference Elicitation Strategies for Supporting Multi-Objective
Decision Making | In multi-objective decision planning and learning, much attention is paid to
producing optimal solution sets that contain an optimal policy for every
possible user preference profile. We argue that the step that follows, i.e,
determining which policy to execute by maximising the user's intrinsic utility
function over this (possibly infinite) set, is under-studied. This paper aims
to fill this gap. We build on previous work on Gaussian processes and pairwise
comparisons for preference modelling, extend it to the multi-objective decision
support scenario, and propose new ordered preference elicitation strategies
based on ranking and clustering. Our main contribution is an in-depth
evaluation of these strategies using computer and human-based experiments. We
show that our proposed elicitation strategies outperform the currently used
pairwise methods, and found that users prefer ranking most. Our experiments
further show that utilising monotonicity information in GPs by using a linear
prior mean at the start and virtual comparisons to the nadir and ideal points,
increases performance. We demonstrate our decision support framework in a
real-world study on traffic regulation, conducted with the city of Amsterdam.
|
Explanations based on the Missing: Towards Contrastive Explanations with
Pertinent Negatives | In this paper we propose a novel method that provides contrastive
explanations justifying the classification of an input by a black box
classifier such as a deep neural network. Given an input we find what should be
%necessarily and minimally and sufficiently present (viz. important object
pixels in an image) to justify its classification and analogously what should
be minimally and necessarily \emph{absent} (viz. certain background pixels). We
argue that such explanations are natural for humans and are used commonly in
domains such as health care and criminology. What is minimally but critically
\emph{absent} is an important part of an explanation, which to the best of our
knowledge, has not been explicitly identified by current explanation methods
that explain predictions of neural networks. We validate our approach on three
real datasets obtained from diverse domains; namely, a handwritten digits
dataset MNIST, a large procurement fraud dataset and a brain activity strength
dataset. In all three cases, we witness the power of our approach in generating
precise explanations that are also easy for human experts to understand and
evaluate.
|
Stochastic Video Generation with a Learned Prior | Generating video frames that accurately predict future world states is
challenging. Existing approaches either fail to capture the full distribution
of outcomes, or yield blurry generations, or both. In this paper we introduce
an unsupervised video generation model that learns a prior model of uncertainty
in a given environment. Video frames are generated by drawing samples from this
prior and combining them with a deterministic estimate of the future frame. The
approach is simple and easily trained end-to-end on a variety of datasets.
Sample generations are both varied and sharp, even many frames into the future,
and compare favorably to those from existing approaches.
|
Approximation Algorithms for Cascading Prediction Models | We present an approximation algorithm that takes a pool of pre-trained models
as input and produces from it a cascaded model with similar accuracy but lower
average-case cost. Applied to state-of-the-art ImageNet classification models,
this yields up to a 2x reduction in floating point multiplications, and up to a
6x reduction in average-case memory I/O. The auto-generated cascades exhibit
intuitive properties, such as using lower-resolution input for easier images
and requiring higher prediction confidence when using a computationally cheaper
model.
|
Detecting Learning vs Memorization in Deep Neural Networks using Shared
Structure Validation Sets | The roles played by learning and memorization represent an important topic in
deep learning research. Recent work on this subject has shown that the
optimization behavior of DNNs trained on shuffled labels is qualitatively
different from DNNs trained with real labels. Here, we propose a novel
permutation approach that can differentiate memorization from learning in deep
neural networks (DNNs) trained as usual (i.e., using the real labels to guide
the learning, rather than shuffled labels). The evaluation of weather the DNN
has learned and/or memorized, happens in a separate step where we compare the
predictive performance of a shallow classifier trained with the features
learned by the DNN, against multiple instances of the same classifier, trained
on the same input, but using shuffled labels as outputs. By evaluating these
shallow classifiers in validation sets that share structure with the training
set, we are able to tell apart learning from memorization. Application of our
permutation approach to multi-layer perceptrons and convolutional neural
networks trained on image data corroborated many findings from other groups.
Most importantly, our illustrations also uncovered interesting dynamic patterns
about how DNNs memorize over increasing numbers of training epochs, and support
the surprising result that DNNs are still able to learn, rather than only
memorize, when trained with pure Gaussian noise as input.
|
Determining the best classifier for predicting the value of a boolean
field on a blood donor database using genetic algorithms | Motivation: Thanks to digitization, we often have access to large databases,
consisting of various fields of information, ranging from numbers to texts and
even boolean values. Such databases lend themselves especially well to machine
learning, classification and big data analysis tasks. We are able to train
classifiers, using already existing data and use them for predicting the values
of a certain field, given that we have information regarding the other fields.
Most specifically, in this study, we look at the Electronic Health Records
(EHRs) that are compiled by hospitals. These EHRs are convenient means of
accessing data of individual patients, but there processing as a whole still
remains a task. However, EHRs that are composed of coherent, well-tabulated
structures lend themselves quite well to the application to machine language,
via the usage of classifiers. In this study, we look at a Blood Transfusion
Service Center Data Set (Data taken from the Blood Transfusion Service Center
in Hsin-Chu City in Taiwan). We used scikit-learn machine learning in python.
From Support Vector Machines(SVM), we use Support Vector Classification(SVC),
from the linear model we import Perceptron. We also used the
K.neighborsclassifier and the decision tree classifiers. Furthermore, we use
the TPOT library to find an optimized pipeline using genetic algorithms. Using
the above classifiers, we score each one of them using k fold cross-validation.
Contact: [email protected] GitHub Repository:
https://github.com/ritabratamaiti/Blooddonorprediction
|
Continuous Relaxation of MAP Inference: A Nonconvex Perspective | In this paper, we study a nonconvex continuous relaxation of MAP inference in
discrete Markov random fields (MRFs). We show that for arbitrary MRFs, this
relaxation is tight, and a discrete stationary point of it can be easily
reached by a simple block coordinate descent algorithm. In addition, we study
the resolution of this relaxation using popular gradient methods, and further
propose a more effective solution using a multilinear decomposition framework
based on the alternating direction method of multipliers (ADMM). Experiments on
many real-world problems demonstrate that the proposed ADMM significantly
outperforms other nonconvex relaxation based methods, and compares favorably
with state of the art MRF optimization algorithms in different settings.
|
Protecting Sensory Data against Sensitive Inferences | There is growing concern about how personal data are used when users grant
applications direct access to the sensors of their mobile devices. In fact,
high resolution temporal data generated by motion sensors reflect directly the
activities of a user and indirectly physical and demographic attributes. In
this paper, we propose a feature learning architecture for mobile devices that
provides flexible and negotiable privacy-preserving sensor data transmission by
appropriately transforming raw sensor data. The objective is to move from the
current binary setting of granting or not permission to an application, toward
a model that allows users to grant each application permission over a limited
range of inferences according to the provided services. The internal structure
of each component of the proposed architecture can be flexibly changed and the
trade-off between privacy and utility can be negotiated between the constraints
of the user and the underlying application. We validated the proposed
architecture in an activity recognition application using two real-world
datasets, with the objective of recognizing an activity without disclosing
gender as an example of private information. Results show that the proposed
framework maintains the usefulness of the transformed data for activity
recognition, with an average loss of only around three percentage points, while
reducing the possibility of gender classification to around 50\%, the target
random guess, from more than 90\% when using raw sensor data. We also present
and distribute MotionSense, a new dataset for activity and attribute
recognition collected from motion sensors.
|
Learning to Explain: An Information-Theoretic Perspective on Model
Interpretation | We introduce instancewise feature selection as a methodology for model
interpretation. Our method is based on learning a function to extract a subset
of features that are most informative for each given example. This feature
selector is trained to maximize the mutual information between selected
features and the response variable, where the conditional distribution of the
response variable given the input is the model to be explained. We develop an
efficient variational approximation to the mutual information, and show the
effectiveness of our method on a variety of synthetic and real data sets using
both quantitative metrics and human evaluation.
|
Variational Inference for Policy Gradient | Inspired by the seminal work on Stein Variational Inference and Stein
Variational Policy Gradient, we derived a method to generate samples from the
posterior variational parameter distribution by \textit{explicitly} minimizing
the KL divergence to match the target distribution in an amortize fashion.
Consequently, we applied this varational inference technique into vanilla
policy gradient, TRPO and PPO with Bayesian Neural Network parameterizations
for reinforcement learning problems.
|
Learning to Gather without Communication | A standard belief on emerging collective behavior is that it emerges from
simple individual rules. Most of the mathematical research on such collective
behavior starts from imperative individual rules, like always go to the center.
But how could an (optimal) individual rule emerge during a short period within
the group lifetime, especially if communication is not available. We argue that
such rules can actually emerge in a group in a short span of time via
collective (multi-agent) reinforcement learning, i.e learning via rewards and
punishments. We consider the gathering problem: several agents (social animals,
swarming robots...) must gather around a same position, which is not determined
in advance. They must do so without communication on their planned decision,
just by looking at the position of other agents. We present the first
experimental evidence that a gathering behavior can be learned without
communication in a partially observable environment. The learned behavior has
the same properties as a self-stabilizing distributed algorithm, as processes
can gather from any initial state (and thus tolerate any transient failure).
Besides, we show that it is possible to tolerate the brutal loss of up to 90\%
of agents without significant impact on the behavior.
|
Federated Meta-Learning with Fast Convergence and Efficient
Communication | Statistical and systematic challenges in collaboratively training machine
learning models across distributed networks of mobile devices have been the
bottlenecks in the real-world application of federated learning. In this work,
we show that meta-learning is a natural choice to handle these issues, and
propose a federated meta-learning framework FedMeta, where a parameterized
algorithm (or meta-learner) is shared, instead of a global model in previous
approaches. We conduct an extensive empirical evaluation on LEAF datasets and a
real-world production dataset, and demonstrate that FedMeta achieves a
reduction in required communication cost by 2.82-4.33 times with faster
convergence, and an increase in accuracy by 3.23%-14.84% as compared to
Federated Averaging (FedAvg) which is a leading optimization algorithm in
federated learning. Moreover, FedMeta preserves user privacy since only the
parameterized algorithm is transmitted between mobile devices and central
servers, and no raw data is collected onto the servers.
|
Pooling homogeneous ensembles to build heterogeneous ones | In ensemble methods, the outputs of a collection of diverse classifiers are
combined in the expectation that the global prediction be more accurate than
the individual ones. Heterogeneous ensembles consist of predictors of different
types, which are likely to have different biases. If these biases are
complementary, the combination of their decisions is beneficial. In this work,
a family of heterogeneous ensembles is built by pooling classifiers from M
homogeneous ensembles of different types of size T. Depending on the fraction
of base classifiers of each type, a particular heterogeneous combination in
this family is represented by a point in a regular simplex in M dimensions. The
M vertices of this simplex represent the different homogeneous ensembles. A
displacement away from one of these vertices effects a smooth transformation of
the corresponding homogeneous ensemble into a heterogeneous one. The optimal
composition of such heterogeneous ensemble can be determined using
cross-validation or, if bootstrap samples are used to build the individual
classifiers, out-of-bag data. An empirical analysis of such combinations of
bootstraped ensembles composed of neural networks, SVMs, and random trees (i.e.
from a standard random forest) illustrates the gains that can be achieved by
this heterogeneous ensemble creation method.
|
Diversity regularization in deep ensembles | Calibrating the confidence of supervised learning models is important for a
variety of contexts where the certainty over predictions should be reliable.
However, it has been reported that deep neural network models are often too
poorly calibrated for achieving complex tasks requiring reliable uncertainty
estimates in their prediction. In this work, we are proposing a strategy for
training deep ensembles with a diversity function regularization, which
improves the calibration property while maintaining a similar prediction
accuracy.
|
Nonlinear Online Learning with Adaptive Nystr\"{o}m Approximation | Use of nonlinear feature maps via kernel approximation has led to success in
many online learning tasks. As a popular kernel approximation method,
Nystr\"{o}m approximation, has been well investigated, and various landmark
points selection methods have been proposed to improve the approximation
quality. However, these improved Nystr\"{o}m methods cannot be directly applied
to the online learning setting as they need to access the entire dataset to
learn the landmark points, while we need to update model on-the-fly in the
online setting. To address this challenge, we propose Adaptive Nystr\"{o}m
approximation for solving nonlinear online learning problems. The key idea is
to adaptively modify the landmark points via online kmeans and adjust the model
accordingly via solving least square problem followed by a gradient descent
step. We show that the resulting algorithm outperforms state-of-the-art online
learning methods under the same budget.
|
Entropy Rate Estimation for Markov Chains with Large State Space | Estimating the entropy based on data is one of the prototypical problems in
distribution property testing and estimation. For estimating the Shannon
entropy of a distribution on $S$ elements with independent samples,
[Paninski2004] showed that the sample complexity is sublinear in $S$, and
[Valiant--Valiant2011] showed that consistent estimation of Shannon entropy is
possible if and only if the sample size $n$ far exceeds $\frac{S}{\log S}$. In
this paper we consider the problem of estimating the entropy rate of a
stationary reversible Markov chain with $S$ states from a sample path of $n$
observations. We show that:
(1) As long as the Markov chain mixes not too slowly, i.e., the relaxation
time is at most $O(\frac{S}{\ln^3 S})$, consistent estimation is achievable
when $n \gg \frac{S^2}{\log S}$.
(2) As long as the Markov chain has some slight dependency, i.e., the
relaxation time is at least $1+\Omega(\frac{\ln^2 S}{\sqrt{S}})$, consistent
estimation is impossible when $n \lesssim \frac{S^2}{\log S}$.
Under both assumptions, the optimal estimation accuracy is shown to be
$\Theta(\frac{S^2}{n \log S})$. In comparison, the empirical entropy rate
requires at least $\Omega(S^2)$ samples to be consistent, even when the Markov
chain is memoryless. In addition to synthetic experiments, we also apply the
estimators that achieve the optimal sample complexity to estimate the entropy
rate of the English language in the Penn Treebank and the Google One Billion
Words corpora, which provides a natural benchmark for language modeling and
relates it directly to the widely used perplexity measure.
|
Learning Mixtures of Linear Regressions with Nearly Optimal Complexity | Mixtures of Linear Regressions (MLR) is an important mixture model with many
applications. In this model, each observation is generated from one of the
several unknown linear regression components, where the identity of the
generated component is also unknown. Previous works either assume strong
assumptions on the data distribution or have high complexity. This paper
proposes a fixed parameter tractable algorithm for the problem under general
conditions, which achieves global convergence and the sample complexity scales
nearly linearly in the dimension. In particular, different from previous works
that require the data to be from the standard Gaussian, the algorithm allows
the data from Gaussians with different covariances. When the conditional number
of the covariances and the number of components are fixed, the algorithm has
nearly optimal sample complexity $N = \tilde{O}(d)$ as well as nearly optimal
computational complexity $\tilde{O}(Nd)$, where $d$ is the dimension of the
data space. To the best of our knowledge, this approach provides the first such
recovery guarantee for this general setting.
|
L2-Nonexpansive Neural Networks | This paper proposes a class of well-conditioned neural networks in which a
unit amount of change in the inputs causes at most a unit amount of change in
the outputs or any of the internal layers. We develop the known methodology of
controlling Lipschitz constants to realize its full potential in maximizing
robustness, with a new regularization scheme for linear layers, new ways to
adapt nonlinearities and a new loss function. With MNIST and CIFAR-10
classifiers, we demonstrate a number of advantages. Without needing any
adversarial training, the proposed classifiers exceed the state of the art in
robustness against white-box L2-bounded adversarial attacks. They generalize
better than ordinary networks from noisy data with partially random labels.
Their outputs are quantitatively meaningful and indicate levels of confidence
and generalization, among other desirable properties.
|
Regional Multi-Armed Bandits | We consider a variant of the classic multi-armed bandit problem where the
expected reward of each arm is a function of an unknown parameter. The arms are
divided into different groups, each of which has a common parameter. Therefore,
when the player selects an arm at each time slot, information of other arms in
the same group is also revealed. This regional bandit model naturally bridges
the non-informative bandit setting where the player can only learn the chosen
arm, and the global bandit model where sampling one arms reveals information of
all arms. We propose an efficient algorithm, UCB-g, that solves the regional
bandit problem by combining the Upper Confidence Bound (UCB) and greedy
principles. Both parameter-dependent and parameter-free regret upper bounds are
derived. We also establish a matching lower bound, which proves the
order-optimality of UCB-g. Moreover, we propose SW-UCB-g, which is an extension
of UCB-g for a non-stationary environment where the parameters slowly vary over
time.
|
The Hidden Vulnerability of Distributed Learning in Byzantium | While machine learning is going through an era of celebrated success,
concerns have been raised about the vulnerability of its backbone: stochastic
gradient descent (SGD). Recent approaches have been proposed to ensure the
robustness of distributed SGD against adversarial (Byzantine) workers sending
poisoned gradients during the training phase. Some of these approaches have
been proven Byzantine-resilient: they ensure the convergence of SGD despite the
presence of a minority of adversarial workers.
We show in this paper that convergence is not enough. In high dimension $d
\gg 1$, an adver\-sary can build on the loss function's non-convexity to make
SGD converge to ineffective models. More precisely, we bring to light that
existing Byzantine-resilient schemes leave a margin of poisoning of
$\Omega\left(f(d)\right)$, where $f(d)$ increases at least like $\sqrt{d~}$.
Based on this leeway, we build a simple attack, and experimentally show its
strong to utmost effectivity on CIFAR-10 and MNIST.
We introduce Bulyan, and prove it significantly reduces the attackers leeway
to a narrow $O( \frac{1}{\sqrt{d~}})$ bound. We empirically show that Bulyan
does not suffer the fragility of existing aggregation rules and, at a
reasonable cost in terms of required batch size, achieves convergence as if
only non-Byzantine gradients had been used to update the model.
|
Asynchronous Byzantine Machine Learning (the case of SGD) | Asynchronous distributed machine learning solutions have proven very
effective so far, but always assuming perfectly functioning workers. In
practice, some of the workers can however exhibit Byzantine behavior, caused by
hardware failures, software bugs, corrupt data, or even malicious attacks. We
introduce \emph{Kardam}, the first distributed asynchronous stochastic gradient
descent (SGD) algorithm that copes with Byzantine workers. Kardam consists of
two complementary components: a filtering and a dampening component. The first
is scalar-based and ensures resilience against $\frac{1}{3}$ Byzantine workers.
Essentially, this filter leverages the Lipschitzness of cost functions and acts
as a self-stabilizer against Byzantine workers that would attempt to corrupt
the progress of SGD. The dampening component bounds the convergence rate by
adjusting to stale information through a generic gradient weighting scheme. We
prove that Kardam guarantees almost sure convergence in the presence of
asynchrony and Byzantine behavior, and we derive its convergence rate. We
evaluate Kardam on the CIFAR-100 and EMNIST datasets and measure its overhead
with respect to non Byzantine-resilient solutions. We empirically show that
Kardam does not introduce additional noise to the learning procedure but does
induce a slowdown (the cost of Byzantine resilience) that we both theoretically
and empirically show to be less than $f/n$, where $f$ is the number of
Byzantine failures tolerated and $n$ the total number of workers.
Interestingly, we also empirically observe that the dampening component is
interesting in its own right for it enables to build an SGD algorithm that
outperforms alternative staleness-aware asynchronous competitors in
environments with honest workers.
|
Actigraphy-based Sleep/Wake Pattern Detection using Convolutional Neural
Networks | Common medical conditions are often associated with sleep abnormalities.
Patients with medical disorders often suffer from poor sleep quality compared
to healthy individuals, which in turn may worsen the symptoms of the disorder.
Accurate detection of sleep/wake patterns is important in developing
personalized digital markers, which can be used for objective measurements and
efficient disease management. Big Data technologies and advanced analytics
methods hold the promise to revolutionize clinical research processes, enabling
the effective blending of digital data into clinical trials. Actigraphy, a
non-invasive activity monitoring method is heavily used to detect and evaluate
activities and movement disorders, and assess sleep/wake behavior. In order to
study the connection between sleep/wake patterns and a cluster headache
disorder, activity data was collected using a wearable device in the course of
a clinical trial. This study presents two novel modeling schemes that utilize
Deep Convolutional Neural Networks (CNN) to identify sleep/wake states. The
proposed methods are a sequential CNN, reminiscent of the bi-directional CNN
for slot filling, and a Multi-Task Learning (MTL) based model. Furthermore, we
expand standard "Sleep" and "Wake" activity states space by adding the "Falling
asleep" and "Siesta" states. We show that the proposed methods provide
promising results in accurate detection of the expanded sleep/wake states.
Finally, we explore the relations between the detected sleep/wake patterns and
onset of cluster headache attacks, and present preliminary observations.
|
The State of the Art in Integrating Machine Learning into Visual
Analytics | Visual analytics systems combine machine learning or other analytic
techniques with interactive data visualization to promote sensemaking and
analytical reasoning. It is through such techniques that people can make sense
of large, complex data. While progress has been made, the tactful combination
of machine learning and data visualization is still under-explored. This
state-of-the-art report presents a summary of the progress that has been made
by highlighting and synthesizing select research advances. Further, it presents
opportunities and challenges to enhance the synergy between machine learning
and visual analytics for impactful future research directions.
|
Incremental and Iterative Learning of Answer Set Programs from Mutually
Distinct Examples | Over the years the Artificial Intelligence (AI) community has produced
several datasets which have given the machine learning algorithms the
opportunity to learn various skills across various domains. However, a subclass
of these machine learning algorithms that aimed at learning logic programs,
namely the Inductive Logic Programming algorithms, have often failed at the
task due to the vastness of these datasets. This has impacted the usability of
knowledge representation and reasoning techniques in the development of AI
systems. In this research, we try to address this scalability issue for the
algorithms that learn answer set programs. We present a sound and complete
algorithm which takes the input in a slightly different manner and performs an
efficient and more user controlled search for a solution. We show via
experiments that our algorithm can learn from two popular datasets from machine
learning community, namely bAbl (a question answering dataset) and MNIST (a
dataset for handwritten digit recognition), which to the best of our knowledge
was not previously possible. The system is publicly available at
https://goo.gl/KdWAcV. This paper is under consideration for acceptance in
TPLP.
|
Robustness of classifiers to uniform $\ell\_p$ and Gaussian noise | We study the robustness of classifiers to various kinds of random noise
models. In particular, we consider noise drawn uniformly from the $\ell\_p$
ball for $p \in [1, \infty]$ and Gaussian noise with an arbitrary covariance
matrix. We characterize this robustness to random noise in terms of the
distance to the decision boundary of the classifier. This analysis applies to
linear classifiers as well as classifiers with locally approximately flat
decision boundaries, a condition which is satisfied by state-of-the-art deep
neural networks. The predicted robustness is verified experimentally.
|
Learning to Route with Sparse Trajectory Sets---Extended Version | Motivated by the increasing availability of vehicle trajectory data, we
propose learn-to-route, a comprehensive trajectory-based routing solution.
Specifically, we first construct a graph-like structure from trajectories as
the routing infrastructure. Second, we enable trajectory-based routing given an
arbitrary (source, destination) pair.
In the first step, given a road network and a collection of trajectories, we
propose a trajectory-based clustering method that identifies regions in a road
network. If a pair of regions are connected by trajectories, we maintain the
paths used by these trajectories and learn a routing preference for travel
between the regions. As trajectories are skewed and sparse, many region pairs
are not connected by trajectories. We thus transfer routing preferences from
region pairs with sufficient trajectories to such region pairs and then use the
transferred preferences to identify paths between the regions. In the second
step, we exploit the above graph-like structure to achieve a comprehensive
trajectory-based routing solution. Empirical studies with two substantial
trajectory data sets offer insight into the proposed solution, indicating that
it is practical. A comparison with a leading routing service offers evidence
that the paper's proposal is able to enhance routing quality.
This is an extended version of "Learning to Route with Sparse Trajectory
Sets" [1], to appear in IEEE ICDE 2018.
|
Sounderfeit: Cloning a Physical Model with Conditional Adversarial
Autoencoders | An adversarial autoencoder conditioned on known parameters of a physical
modeling bowed string synthesizer is evaluated for use in parameter estimation
and resynthesis tasks. Latent dimensions are provided to capture variance not
explained by the conditional parameters. Results are compared with and without
the adversarial training, and a system capable of "copying" a given
parameter-signal bidirectional relationship is examined. A real-time synthesis
system built on a generative, conditioned and regularized neural network is
presented, allowing to construct engaging sound synthesizers based purely on
recorded data.
|
Iterate averaging as regularization for stochastic gradient descent | We propose and analyze a variant of the classic Polyak-Ruppert averaging
scheme, broadly used in stochastic gradient methods. Rather than a uniform
average of the iterates, we consider a weighted average, with weights decaying
in a geometric fashion. In the context of linear least squares regression, we
show that this averaging scheme has a the same regularizing effect, and indeed
is asymptotically equivalent, to ridge regression. In particular, we derive
finite-sample bounds for the proposed approach that match the best known
results for regularized stochastic gradient methods.
|
Learning Topic Models by Neighborhood Aggregation | Topic models are frequently used in machine learning owing to their high
interpretability and modular structure. However, extending a topic model to
include a supervisory signal, to incorporate pre-trained word embedding vectors
and to include a nonlinear output function is not an easy task because one has
to resort to a highly intricate approximate inference procedure. The present
paper shows that topic modeling with pre-trained word embedding vectors can be
viewed as implementing a neighborhood aggregation algorithm where messages are
passed through a network defined over words. From the network view of topic
models, nodes correspond to words in a document and edges correspond to either
a relationship describing co-occurring words in a document or a relationship
describing the same word in the corpus. The network view allows us to extend
the model to include supervisory signals, incorporate pre-trained word
embedding vectors and include a nonlinear output function in a simple manner.
In experiments, we show that our approach outperforms the state-of-the-art
supervised Latent Dirichlet Allocation implementation in terms of held-out
document classification tasks.
|
Intrinsic Motivation and Mental Replay enable Efficient Online
Adaptation in Stochastic Recurrent Networks | Autonomous robots need to interact with unknown, unstructured and changing
environments, constantly facing novel challenges. Therefore, continuous online
adaptation for lifelong-learning and the need of sample-efficient mechanisms to
adapt to changes in the environment, the constraints, the tasks, or the robot
itself are crucial. In this work, we propose a novel framework for
probabilistic online motion planning with online adaptation based on a
bio-inspired stochastic recurrent neural network. By using learning signals
which mimic the intrinsic motivation signalcognitive dissonance in addition
with a mental replay strategy to intensify experiences, the stochastic
recurrent network can learn from few physical interactions and adapts to novel
environments in seconds. We evaluate our online planning and adaptation
framework on an anthropomorphic KUKA LWR arm. The rapid online adaptation is
shown by learning unknown workspace constraints sample-efficiently from few
physical interactions while following given way points.
|
VBALD - Variational Bayesian Approximation of Log Determinants | Evaluating the log determinant of a positive definite matrix is ubiquitous in
machine learning. Applications thereof range from Gaussian processes,
minimum-volume ellipsoids, metric learning, kernel learning, Bayesian neural
networks, Determinental Point Processes, Markov random fields to partition
functions of discrete graphical models. In order to avoid the canonical, yet
prohibitive, Cholesky $\mathcal{O}(n^{3})$ computational cost, we propose a
novel approach, with complexity $\mathcal{O}(n^{2})$, based on a constrained
variational Bayes algorithm. We compare our method to Taylor, Chebyshev and
Lanczos approaches and show state of the art performance on both synthetic and
real-world datasets.
|
Sampling as optimization in the space of measures: The Langevin dynamics
as a composite optimization problem | We study sampling as optimization in the space of measures. We focus on
gradient flow-based optimization with the Langevin dynamics as a case study. We
investigate the source of the bias of the unadjusted Langevin algorithm (ULA)
in discrete time, and consider how to remove or reduce the bias. We point out
the difficulty is that the heat flow is exactly solvable, but neither its
forward nor backward method is implementable in general, except for Gaussian
data. We propose the symmetrized Langevin algorithm (SLA), which should have a
smaller bias than ULA, at the price of implementing a proximal gradient step in
space. We show SLA is in fact consistent for Gaussian target measure, whereas
ULA is not. We also illustrate various algorithms explicitly for Gaussian
target measure, including gradient descent, proximal gradient, and
Forward-Backward, and show they are all consistent.
|
Sliding Bidirectional Recurrent Neural Networks for Sequence Detection
in Communication Systems | The design and analysis of communication systems typically rely on the
development of mathematical models that describe the underlying communication
channel. However, in some systems, such as molecular communication systems
where chemical signals are used for transfer of information, the underlying
channel models are unknown. In these scenarios, a completely new approach to
design and analysis is required. In this work, we focus on one important aspect
of communication systems, the detection algorithms, and demonstrate that by
using tools from deep learning, it is possible to train detectors that perform
well without any knowledge of the underlying channel models. We propose a
technique we call sliding bidirectional recurrent neural network (SBRNN) for
real-time sequence detection. We evaluate this algorithm using experimental
data that is collected by a chemical communication platform, where the channel
model is unknown and difficult to model analytically. We show that deep
learning algorithms perform significantly better than a detector proposed in
previous works, and the SBRNN outperforms other techniques considered in this
work.
|
Collaboratively Learning the Best Option, Using Bounded Memory | We consider multi-armed bandit problems in social groups wherein each
individual has bounded memory and shares the common goal of learning the best
arm/option. We say an individual learns the best option if eventually (as $t
\to \infty$) it pulls only the arm with the highest average reward. While this
goal is provably impossible for an isolated individual, we show that, in social
groups, this goal can be achieved easily with the aid of social persuasion,
i.e., communication. Specifically, we study the learning dynamics wherein an
individual sequentially decides on which arm to pull next based on not only its
private reward feedback but also the suggestions provided by randomly chosen
peers. Our learning dynamics are hard to analyze via explicit probabilistic
calculations due to the stochastic dependency induced by social interaction.
Instead, we employ the mean-field approximation method from statistical physics
and we show:
(1) With probability $\to 1$ as the social group size $N \to \infty $, every
individual in the social group learns the best option.
(2) Over an arbitrary finite time horizon $[0, T]$, with high probability (in
$N$), the fraction of individuals that prefer the best option grows to 1
exponentially fast as $t$ increases ($t\in [0, T]$).
A major innovation of our mean-filed analysis is a simple yet powerful
technique to deal with absorbing states in the interchange of limits $N \to
\infty$ and $t \to \infty $. The mean-field approximation method allows us to
approximate the probabilistic sample paths of our learning dynamics by a
deterministic and smooth trajectory that corresponds to the unique solution of
a well-behaved system of ordinary differential equations (ODEs). Such an
approximation is desired because the analysis of a system of ODEs is relatively
easier than that of the original stochastic system.
|
Projection-Free Online Optimization with Stochastic Gradient: From
Convexity to Submodularity | Online optimization has been a successful framework for solving large-scale
problems under computational constraints and partial information. Current
methods for online convex optimization require either a projection or exact
gradient computation at each step, both of which can be prohibitively expensive
for large-scale applications. At the same time, there is a growing trend of
non-convex optimization in machine learning community and a need for online
methods. Continuous DR-submodular functions, which exhibit a natural
diminishing returns condition, have recently been proposed as a broad class of
non-convex functions which may be efficiently optimized. Although online
methods have been introduced, they suffer from similar problems. In this work,
we propose Meta-Frank-Wolfe, the first online projection-free algorithm that
uses stochastic gradient estimates. The algorithm relies on a careful sampling
of gradients in each round and achieves the optimal $O( \sqrt{T})$ adversarial
regret bounds for convex and continuous submodular optimization. We also
propose One-Shot Frank-Wolfe, a simpler algorithm which requires only a single
stochastic gradient estimate in each round and achieves an $O(T^{2/3})$
stochastic regret bound for convex and continuous submodular optimization. We
apply our methods to develop a novel "lifting" framework for the online
discrete submodular maximization and also see that they outperform current
state-of-the-art techniques on various experiments.
|
Adversarial Examples that Fool both Computer Vision and Time-Limited
Humans | Machine learning models are vulnerable to adversarial examples: small changes
to images can cause computer vision models to make mistakes such as identifying
a school bus as an ostrich. However, it is still an open question whether
humans are prone to similar mistakes. Here, we address this question by
leveraging recent techniques that transfer adversarial examples from computer
vision models with known parameters and architecture to other models with
unknown parameters and architecture, and by matching the initial processing of
the human visual system. We find that adversarial examples that strongly
transfer across computer vision models influence the classifications made by
time-limited human observers.
|
Tensor field networks: Rotation- and translation-equivariant neural
networks for 3D point clouds | We introduce tensor field neural networks, which are locally equivariant to
3D rotations, translations, and permutations of points at every layer. 3D
rotation equivariance removes the need for data augmentation to identify
features in arbitrary orientations. Our network uses filters built from
spherical harmonics; due to the mathematical consequences of this filter
choice, each layer accepts as input (and guarantees as output) scalars,
vectors, and higher-order tensors, in the geometric sense of these terms. We
demonstrate the capabilities of tensor field networks with tasks in geometry,
physics, and chemistry.
|
The Secret Sharer: Evaluating and Testing Unintended Memorization in
Neural Networks | This paper describes a testing methodology for quantitatively assessing the
risk that rare or unique training-data sequences are unintentionally memorized
by generative sequence models---a common type of machine-learning model.
Because such models are sometimes trained on sensitive data (e.g., the text of
users' private messages), this methodology can benefit privacy by allowing
deep-learning practitioners to select means of training that minimize such
memorization.
In experiments, we show that unintended memorization is a persistent,
hard-to-avoid issue that can have serious consequences. Specifically, for
models trained without consideration of memorization, we describe new,
efficient procedures that can extract unique, secret sequences, such as credit
card numbers. We show that our testing strategy is a practical and easy-to-use
first line of defense, e.g., by describing its application to quantitatively
limit data exposure in Google's Smart Compose, a commercial text-completion
neural network trained on millions of users' email messages.
|
Vector Field Based Neural Networks | A novel Neural Network architecture is proposed using the mathematically and
physically rich idea of vector fields as hidden layers to perform nonlinear
transformations in the data. The data points are interpreted as particles
moving along a flow defined by the vector field which intuitively represents
the desired movement to enable classification. The architecture moves the data
points from their original configuration to anew one following the streamlines
of the vector field with the objective of achieving a final configuration where
classes are separable. An optimization problem is solved through gradient
descent to learn this vector field.
|
Hessian-based Analysis of Large Batch Training and Robustness to
Adversaries | Large batch size training of Neural Networks has been shown to incur accuracy
loss when trained with the current methods. The exact underlying reasons for
this are still not completely understood. Here, we study large batch size
training through the lens of the Hessian operator and robust optimization. In
particular, we perform a Hessian based study to analyze exactly how the
landscape of the loss function changes when training with large batch size. We
compute the true Hessian spectrum, without approximation, by back-propagating
the second derivative. Extensive experiments on multiple networks show that
saddle-points are not the cause for generalization gap of large batch size
training, and the results consistently show that large batch converges to
points with noticeably higher Hessian spectrum. Furthermore, we show that
robust training allows one to favor flat areas, as points with large Hessian
spectrum show poor robustness to adversarial perturbation. We further study
this relationship, and provide empirical and theoretical proof that the inner
loop for robust training is a saddle-free optimization problem \textit{almost
everywhere}. We present detailed experiments with five different network
architectures, including a residual network, tested on MNIST, CIFAR-10, and
CIFAR-100 datasets. We have open sourced our method which can be accessed at
[1].
|
Characterizing Implicit Bias in Terms of Optimization Geometry | We study the implicit bias of generic optimization methods, such as mirror
descent, natural gradient descent, and steepest descent with respect to
different potentials and norms, when optimizing underdetermined linear
regression or separable linear classification problems. We explore the question
of whether the specific global minimum (among the many possible global minima)
reached by an algorithm can be characterized in terms of the potential or norm
of the optimization geometry, and independently of hyperparameter choices such
as step-size and momentum.
|
On the Convergence and Robustness of Training GANs with Regularized
Optimal Transport | Generative Adversarial Networks (GANs) are one of the most practical methods
for learning data distributions. A popular GAN formulation is based on the use
of Wasserstein distance as a metric between probability distributions.
Unfortunately, minimizing the Wasserstein distance between the data
distribution and the generative model distribution is a computationally
challenging problem as its objective is non-convex, non-smooth, and even hard
to compute. In this work, we show that obtaining gradient information of the
smoothed Wasserstein GAN formulation, which is based on regularized Optimal
Transport (OT), is computationally effortless and hence one can apply first
order optimization methods to minimize this objective. Consequently, we
establish theoretical convergence guarantee to stationarity for a proposed
class of GAN optimization algorithms. Unlike the original non-smooth
formulation, our algorithm only requires solving the discriminator to
approximate optimality. We apply our method to learning MNIST digits as well as
CIFAR-10images. Our experiments show that our method is computationally
efficient and generates images comparable to the state of the art algorithms
given the same architecture and computational power.
|
SeNA-CNN: Overcoming Catastrophic Forgetting in Convolutional Neural
Networks by Selective Network Augmentation | Lifelong learning aims to develop machine learning systems that can learn new
tasks while preserving the performance on previous learned tasks. In this paper
we present a method to overcome catastrophic forgetting on convolutional neural
networks, that learns new tasks and preserves the performance on old tasks
without accessing the data of the original model, by selective network
augmentation. The experiment results showed that SeNA-CNN, in some scenarios,
outperforms the state-of-art Learning without Forgetting algorithm. Results
also showed that in some situations it is better to use SeNA-CNN instead of
training a neural network using isolated learning.
|
Confidential Boosting with Random Linear Classifiers for Outsourced
User-generated Data | User-generated data is crucial to predictive modeling in many applications.
With a web/mobile/wearable interface, a data owner can continuously record data
generated by distributed users and build various predictive models from the
data to improve their operations, services, and revenue. Due to the large size
and evolving nature of users data, data owners may rely on public cloud service
providers (Cloud) for storage and computation scalability. Exposing sensitive
user-generated data and advanced analytic models to Cloud raises privacy
concerns. We present a confidential learning framework, SecureBoost, for data
owners that want to learn predictive models from aggregated user-generated data
but offload the storage and computational burden to Cloud without having to
worry about protecting the sensitive data. SecureBoost allows users to submit
encrypted or randomly masked data to designated Cloud directly. Our framework
utilizes random linear classifiers (RLCs) as the base classifiers in the
boosting framework to dramatically simplify the design of the proposed
confidential boosting protocols, yet still preserve the model quality. A
Cryptographic Service Provider (CSP) is used to assist the Cloud's processing,
reducing the complexity of the protocol constructions. We present two
constructions of SecureBoost: HE+GC and SecSh+GC, using combinations of
homomorphic encryption, garbled circuits, and random masking to achieve both
security and efficiency. For a boosted model, Cloud learns only the RLCs and
the CSP learns only the weights of the RLCs. Finally, the data owner collects
the two parts to get the complete model. We conduct extensive experiments to
understand the quality of the RLC-based boosting and the cost distribution of
the constructions. Our results show that SecureBoost can efficiently learn
high-quality boosting models from protected user-generated data.
|
Locally Adaptive Learning Loss for Semantic Image Segmentation | We propose a novel locally adaptive learning estimator for enhancing the
inter- and intra- discriminative capabilities of Deep Neural Networks, which
can be used as improved loss layer for semantic image segmentation tasks. Most
loss layers compute pixel-wise cost between feature maps and ground truths,
ignoring spatial layouts and interactions between neighboring pixels with same
object category, and thus networks cannot be effectively sensitive to
intra-class connections. Stride by stride, our method firstly conducts adaptive
pooling filter operating over predicted feature maps, aiming to merge predicted
distributions over a small group of neighboring pixels with same category, and
then it computes cost between the merged distribution vector and their category
label. Such design can make groups of neighboring predictions from same
category involved into estimations on predicting correctness with respect to
their category, and hence train networks to be more sensitive to regional
connections between adjacent pixels based on their categories. In the
experiments on Pascal VOC 2012 segmentation datasets, the consistently improved
results show that our proposed approach achieves better segmentation masks
against previous counterparts.
|
Unicorn: Continual Learning with a Universal, Off-policy Agent | Some real-world domains are best characterized as a single task, but for
others this perspective is limiting. Instead, some tasks continually grow in
complexity, in tandem with the agent's competence. In continual learning, also
referred to as lifelong learning, there are no explicit task boundaries or
curricula. As learning agents have become more powerful, continual learning
remains one of the frontiers that has resisted quick progress. To test
continual learning capabilities we consider a challenging 3D domain with an
implicit sequence of tasks and sparse rewards. We propose a novel agent
architecture called Unicorn, which demonstrates strong continual learning and
outperforms several baseline agents on the proposed domain. The agent achieves
this by jointly representing and learning multiple policies efficiently, using
a parallel off-policy learning setup.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.