title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
One-Class Adversarial Nets for Fraud Detection | Many online applications, such as online social networks or knowledge bases,
are often attacked by malicious users who commit different types of actions
such as vandalism on Wikipedia or fraudulent reviews on eBay. Currently, most
of the fraud detection approaches require a training dataset that contains
records of both benign and malicious users. However, in practice, there are
often no or very few records of malicious users. In this paper, we develop
one-class adversarial nets (OCAN) for fraud detection using training data with
only benign users. OCAN first uses LSTM-Autoencoder to learn the
representations of benign users from their sequences of online activities. It
then detects malicious users by training a discriminator with a complementary
GAN model that is different from the regular GAN model. Experimental results
show that our OCAN outperforms the state-of-the-art one-class classification
models and achieves comparable performance with the latest multi-source LSTM
model that requires both benign and malicious users in the training phase.
|
Norm matters: efficient and accurate normalization schemes in deep
networks | Over the past few years, Batch-Normalization has been commonly used in deep
networks, allowing faster training and high performance for a wide variety of
applications. However, the reasons behind its merits remained unanswered, with
several shortcomings that hindered its use for certain tasks. In this work, we
present a novel view on the purpose and function of normalization methods and
weight-decay, as tools to decouple weights' norm from the underlying optimized
objective. This property highlights the connection between practices such as
normalization, weight decay and learning-rate adjustments. We suggest several
alternatives to the widely used $L^2$ batch-norm, using normalization in $L^1$
and $L^\infty$ spaces that can substantially improve numerical stability in
low-precision implementations as well as provide computational and memory
benefits. We demonstrate that such methods enable the first batch-norm
alternative to work for half-precision implementations. Finally, we suggest a
modification to weight-normalization, which improves its performance on
large-scale tasks.
|
Marginal Singularity, and the Benefits of Labels in Covariate-Shift | We present new minimax results that concisely capture the relative benefits
of source and target labeled data, under covariate-shift. Namely, we show that
the benefits of target labels are controlled by a transfer-exponent $\gamma$
that encodes how singular Q is locally w.r.t. P, and interestingly allows
situations where transfer did not seem possible under previous insights. In
fact, our new minimax analysis - in terms of $\gamma$ - reveals a continuum of
regimes ranging from situations where target labels have little benefit, to
regimes where target labels dramatically improve classification. We then show
that a recently proposed semi-supervised procedure can be extended to adapt to
unknown $\gamma$, and therefore requests labels only when beneficial, while
achieving minimax transfer rates.
|
Conducting Credit Assignment by Aligning Local Representations | Using back-propagation and its variants to train deep networks is often
problematic for new users. Issues such as exploding gradients, vanishing
gradients, and high sensitivity to weight initialization strategies often make
networks difficult to train, especially when users are experimenting with new
architectures. Here, we present Local Representation Alignment (LRA), a
training procedure that is much less sensitive to bad initializations, does not
require modifications to the network architecture, and can be adapted to
networks with highly nonlinear and discrete-valued activation functions.
Furthermore, we show that one variation of LRA can start with a null
initialization of network weights and still successfully train networks with a
wide variety of nonlinearities, including tanh, ReLU-6, softplus, signum and
others that may draw their inspiration from biology.
A comprehensive set of experiments on MNIST and the much harder Fashion MNIST
data sets show that LRA can be used to train networks robustly and effectively,
succeeding even when back-propagation fails and outperforming other alternative
learning algorithms, such as target propagation and feedback alignment.
|
ST-GAN: Spatial Transformer Generative Adversarial Networks for Image
Compositing | We address the problem of finding realistic geometric corrections to a
foreground object such that it appears natural when composited into a
background image. To achieve this, we propose a novel Generative Adversarial
Network (GAN) architecture that utilizes Spatial Transformer Networks (STNs) as
the generator, which we call Spatial Transformer GANs (ST-GANs). ST-GANs seek
image realism by operating in the geometric warp parameter space. In
particular, we exploit an iterative STN warping scheme and propose a sequential
training strategy that achieves better results compared to naive training of a
single generator. One of the key advantages of ST-GAN is its applicability to
high-resolution images indirectly since the predicted warp parameters are
transferable between reference frames. We demonstrate our approach in two
applications: (1) visualizing how indoor furniture (e.g. from product images)
might be perceived in a room, (2) hallucinating how accessories like glasses
would look when matched with real portraits.
|
TACO: Learning Task Decomposition via Temporal Alignment for Control | Many advanced Learning from Demonstration (LfD) methods consider the
decomposition of complex, real-world tasks into simpler sub-tasks. By reusing
the corresponding sub-policies within and between tasks, they provide training
data for each policy from different high-level tasks and compose them to
perform novel ones. Existing approaches to modular LfD focus either on learning
a single high-level task or depend on domain knowledge and temporal
segmentation. In contrast, we propose a weakly supervised, domain-agnostic
approach based on task sketches, which include only the sequence of sub-tasks
performed in each demonstration. Our approach simultaneously aligns the
sketches with the observed demonstrations and learns the required sub-policies.
This improves generalisation in comparison to separate optimisation procedures.
We evaluate the approach on multiple domains, including a simulated 3D robot
arm control task using purely image-based observations. The results show that
our approach performs commensurately with fully supervised approaches, while
requiring significantly less annotation effort.
|
Pathological Analysis of Stress Urinary Incontinence in Females using
Artificial Neural Networks | Objectives: To mathematically investigate urethral pressure and influencing
parameters of stress urinary incontinence (SUI) in women, with focus on the
clinical aspects of the mathematical modeling.
Method: Several patients' data are extracted from UPP and urodynamic
documents and their relation and affinities are modeled using an artificial
neural network (ANN) model. The studied parameter is urethral pressure as a
function of two variables: the age of the patient and the position in which the
pressure was measured across the urethra (normalized length).
Results: The ANN-generated surface, showing the relation between the chosen
parameters and the urethral pressure in the studied patients, is more efficient
than the surface generated by conventional mathematical methods for clinical
analysis, with multi-sample analysis being obtained. For example, in elderly
people, there are many low-pressure zones throughout the urethra length,
indicating that there is more incontinence in old age.
Conclusion: The predictions of urethral pressure made by the trained neural
network model in relation to the studied effective parameters can be used to
build a medical assistance system in order to help clinicians diagnose urinary
incontinence problems more efficiently.
|
AspEm: Embedding Learning by Aspects in Heterogeneous Information
Networks | Heterogeneous information networks (HINs) are ubiquitous in real-world
applications. Due to the heterogeneity in HINs, the typed edges may not fully
align with each other. In order to capture the semantic subtlety, we propose
the concept of aspects with each aspect being a unit representing one
underlying semantic facet. Meanwhile, network embedding has emerged as a
powerful method for learning network representation, where the learned
embedding can be used as features in various downstream applications.
Therefore, we are motivated to propose a novel embedding learning
framework---AspEm---to preserve the semantic information in HINs based on
multiple aspects. Instead of preserving information of the network in one
semantic space, AspEm encapsulates information regarding each aspect
individually. In order to select aspects for embedding purpose, we further
devise a solution for AspEm based on dataset-wide statistics. To corroborate
the efficacy of AspEm, we conducted experiments on two real-words datasets with
two types of applications---classification and link prediction. Experiment
results demonstrate that AspEm can outperform baseline network embedding
learning methods by considering multiple aspects, where the aspects can be
selected from the given HIN in an unsupervised manner.
|
On Discrimination Discovery and Removal in Ranked Data using Causal
Graph | Predictive models learned from historical data are widely used to help
companies and organizations make decisions. However, they may digitally
unfairly treat unwanted groups, raising concerns about fairness and
discrimination. In this paper, we study the fairness-aware ranking problem
which aims to discover discrimination in ranked datasets and reconstruct the
fair ranking. Existing methods in fairness-aware ranking are mainly based on
statistical parity that cannot measure the true discriminatory effect since
discrimination is causal. On the other hand, existing methods in causal-based
anti-discrimination learning focus on classification problems and cannot be
directly applied to handle the ranked data. To address these limitations, we
propose to map the rank position to a continuous score variable that represents
the qualification of the candidates. Then, we build a causal graph that
consists of both the discrete profile attributes and the continuous score. The
path-specific effect technique is extended to the mixed-variable causal graph
to identify both direct and indirect discrimination. The relationship between
the path-specific effects for the ranked data and those for the binary decision
is theoretically analyzed. Finally, algorithms for discovering and removing
discrimination from a ranked dataset are developed. Experiments using the real
dataset show the effectiveness of our approaches.
|
Convergence of Gradient Descent on Separable Data | We provide a detailed study on the implicit bias of gradient descent when
optimizing loss functions with strictly monotone tails, such as the logistic
loss, over separable datasets. We look at two basic questions: (a) what are the
conditions on the tail of the loss function under which gradient descent
converges in the direction of the $L_2$ maximum-margin separator? (b) how does
the rate of margin convergence depend on the tail of the loss function and the
choice of the step size? We show that for a large family of super-polynomial
tailed losses, gradient descent iterates on linear networks of any depth
converge in the direction of $L_2$ maximum-margin solution, while this does not
hold for losses with heavier tails. Within this family, for simple linear
models we show that the optimal rates with fixed step size is indeed obtained
for the commonly used exponentially tailed losses such as logistic loss.
However, with a fixed step size the optimal convergence rate is extremely slow
as $1/\log(t)$, as also proved in Soudry et al. (2018). For linear models with
exponential loss, we further prove that the convergence rate could be improved
to $\log (t) /\sqrt{t}$ by using aggressive step sizes that compensates for the
rapidly vanishing gradients. Numerical results suggest this method might be
useful for deep networks.
|
Energy-entropy competition and the effectiveness of stochastic gradient
descent in machine learning | Finding parameters that minimise a loss function is at the core of many
machine learning methods. The Stochastic Gradient Descent algorithm is widely
used and delivers state of the art results for many problems. Nonetheless,
Stochastic Gradient Descent typically cannot find the global minimum, thus its
empirical effectiveness is hitherto mysterious. We derive a correspondence
between parameter inference and free energy minimisation in statistical
physics. The degree of undersampling plays the role of temperature. Analogous
to the energy-entropy competition in statistical physics, wide but shallow
minima can be optimal if the system is undersampled, as is typical in many
applications. Moreover, we show that the stochasticity in the algorithm has a
non-trivial correlation structure which systematically biases it towards wide
minima. We illustrate our argument with two prototypical models: image
classification using deep learning, and a linear neural network where we can
analytically reveal the relationship between entropy and out-of-sample error.
|
Thermodynamics of Restricted Boltzmann Machines and related learning
dynamics | We investigate the thermodynamic properties of a Restricted Boltzmann Machine
(RBM), a simple energy-based generative model used in the context of
unsupervised learning. Assuming the information content of this model to be
mainly reflected by the spectral properties of its weight matrix $W$, we try to
make a realistic analysis by averaging over an appropriate statistical ensemble
of RBMs.
First, a phase diagram is derived. Otherwise similar to that of the
Sherrington- Kirkpatrick (SK) model with ferromagnetic couplings, the RBM's
phase diagram presents a ferromagnetic phase which may or may not be of
compositional type depending on the kurtosis of the distribution of the
components of the singular vectors of $W$.
Subsequently, the learning dynamics of the RBM is studied in the
thermodynamic limit. A "typical" learning trajectory is shown to solve an
effective dynamical equation, based on the aforementioned ensemble average and
explicitly involving order parameters obtained from the thermodynamic analysis.
In particular, this let us show how the evolution of the dominant singular
values of $W$, and thus of the unstable modes, is driven by the input data. At
the beginning of the training, in which the RBM is found to operate in the
linear regime, the unstable modes reflect the dominant covariance modes of the
data. In the non-linear regime, instead, the selected modes interact and
eventually impose a matching of the order parameters to their empirical
counterparts estimated from the data.
Finally, we illustrate our considerations by performing experiments on both
artificial and real data, showing in particular how the RBM operates in the
ferromagnetic compositional phase.
|
An Online Algorithm for Learning Buyer Behavior under Realistic Pricing
Restrictions | We propose a new efficient online algorithm to learn the parameters governing
the purchasing behavior of a utility maximizing buyer, who responds to prices,
in a repeated interaction setting. The key feature of our algorithm is that it
can learn even non-linear buyer utility while working with arbitrary price
constraints that the seller may impose. This overcomes a major shortcoming of
previous approaches, which use unrealistic prices to learn these parameters
making them unsuitable in practice.
|
Learning Filter Bank Sparsifying Transforms | Data is said to follow the transform (or analysis) sparsity model if it
becomes sparse when acted on by a linear operator called a sparsifying
transform. Several algorithms have been designed to learn such a transform
directly from data, and data-adaptive sparsifying transforms have demonstrated
excellent performance in signal restoration tasks. Sparsifying transforms are
typically learned using small sub-regions of data called patches, but these
algorithms often ignore redundant information shared between neighboring
patches.
We show that many existing transform and analysis sparse representations can
be viewed as filter banks, thus linking the local properties of patch-based
model to the global properties of a convolutional model. We propose a new
transform learning framework where the sparsifying transform is an undecimated
perfect reconstruction filter bank. Unlike previous transform learning
algorithms, the filter length can be chosen independently of the number of
filter bank channels. Numerical results indicate filter bank sparsifying
transforms outperform existing patch-based transform learning for image
denoising while benefiting from additional flexibility in the design process.
|
Occupancy Map Prediction Using Generative and Fully Convolutional
Networks for Vehicle Navigation | Fast, collision-free motion through unknown environments remains a
challenging problem for robotic systems. In these situations, the robot's
ability to reason about its future motion is often severely limited by sensor
field of view (FOV). By contrast, biological systems routinely make decisions
by taking into consideration what might exist beyond their FOV based on prior
experience. In this paper, we present an approach for predicting occupancy map
representations of sensor data for future robot motions using deep neural
networks. We evaluate several deep network architectures, including purely
generative and adversarial models. Testing on both simulated and real
environments we demonstrated performance both qualitatively and quantitatively,
with SSIM similarity measure up to 0.899. We showed that it is possible to make
predictions about occupied space beyond the physical robot's FOV from simulated
training data. In the future, this method will allow robots to navigate through
unknown environments in a faster, safer manner.
|
Understanding Short-Horizon Bias in Stochastic Meta-Optimization | Careful tuning of the learning rate, or even schedules thereof, can be
crucial to effective neural net training. There has been much recent interest
in gradient-based meta-optimization, where one tunes hyperparameters, or even
learns an optimizer, in order to minimize the expected loss when the training
procedure is unrolled. But because the training procedure must be unrolled
thousands of times, the meta-objective must be defined with an
orders-of-magnitude shorter time horizon than is typical for neural net
training. We show that such short-horizon meta-objectives cause a serious bias
towards small step sizes, an effect we term short-horizon bias. We introduce a
toy problem, a noisy quadratic cost function, on which we analyze short-horizon
bias by deriving and comparing the optimal schedules for short and long time
horizons. We then run meta-optimization experiments (both offline and online)
on standard benchmark datasets, showing that meta-optimization chooses too
small a learning rate by multiple orders of magnitude, even when run with a
moderately long time horizon (100 steps) typical of work in the area. We
believe short-horizon bias is a fundamental problem that needs to be addressed
if meta-optimization is to scale to practical neural net training regimes.
|
Accelerated Gradient Boosting | Gradient tree boosting is a prediction algorithm that sequentially produces a
model in the form of linear combinations of decision trees, by solving an
infinite-dimensional optimization problem. We combine gradient boosting and
Nesterov's accelerated descent to design a new algorithm, which we call AGB
(for Accelerated Gradient Boosting). Substantial numerical evidence is provided
on both synthetic and real-life data sets to assess the excellent performance
of the method in a large variety of prediction problems. It is empirically
shown that AGB is much less sensitive to the shrinkage parameter and outputs
predictors that are considerably more sparse in the number of trees, while
retaining the exceptional performance of gradient boosting.
|
Online Deep Learning: Growing RBM on the fly | We propose a novel online learning algorithm for Restricted Boltzmann
Machines (RBM), namely, the Online Generative Discriminative Restricted
Boltzmann Machine (OGD-RBM), that provides the ability to build and adapt the
network architecture of RBM according to the statistics of streaming data. The
OGD-RBM is trained in two phases: (1) an online generative phase for
unsupervised feature representation at the hidden layer and (2) a
discriminative phase for classification. The online generative training begins
with zero neurons in the hidden layer, adds and updates the neurons to adapt to
statistics of streaming data in a single pass unsupervised manner, resulting in
a feature representation best suited to the data. The discriminative phase is
based on stochastic gradient descent and associates the represented features to
the class labels. We demonstrate the OGD-RBM on a set of multi-category and
binary classification problems for data sets having varying degrees of
class-imbalance. We first apply the OGD-RBM algorithm on the multi-class MNIST
dataset to characterize the network evolution. We demonstrate that the online
generative phase converges to a stable, concise network architecture, wherein
individual neurons are inherently discriminative to the class labels despite
unsupervised training. We then benchmark OGD-RBM performance to other machine
learning, neural network and ClassRBM techniques for credit scoring
applications using 3 public non-stationary two-class credit datasets with
varying degrees of class-imbalance. We report that OGD-RBM improves accuracy by
2.5-3% over batch learning techniques while requiring at least 24%-70% fewer
neurons and fewer training samples. This online generative training approach
can be extended greedily to multiple layers for training Deep Belief Networks
in non-stationary data mining applications without the need for a priori fixed
architectures.
|
The Earth ain't Flat: Monocular Reconstruction of Vehicles on Steep and
Graded Roads from a Moving Camera | Accurate localization of other traffic participants is a vital task in
autonomous driving systems. State-of-the-art systems employ a combination of
sensing modalities such as RGB cameras and LiDARs for localizing traffic
participants, but most such demonstrations have been confined to plain roads.
We demonstrate, to the best of our knowledge, the first results for monocular
object localization and shape estimation on surfaces that do not share the same
plane with the moving monocular camera. We approximate road surfaces by local
planar patches and use semantic cues from vehicles in the scene to initialize a
local bundle-adjustment like procedure that simultaneously estimates the pose
and shape of the vehicles, and the orientation of the local ground plane on
which the vehicle stands as well. We evaluate the proposed approach on the
KITTI and SYNTHIA-SF benchmarks, for a variety of road plane configurations.
The proposed approach significantly improves the state-of-the-art for monocular
object localization on arbitrarily-shaped roads.
|
The Contextual Loss for Image Transformation with Non-Aligned Data | Feed-forward CNNs trained for image transformation problems rely on loss
functions that measure the similarity between the generated image and a target
image. Most of the common loss functions assume that these images are spatially
aligned and compare pixels at corresponding locations. However, for many tasks,
aligned training pairs of images will not be available. We present an
alternative loss function that does not require alignment, thus providing an
effective and simple solution for a new space of problems. Our loss is based on
both context and semantics -- it compares regions with similar semantic
meaning, while considering the context of the entire image. Hence, for example,
when transferring the style of one face to another, it will translate
eyes-to-eyes and mouth-to-mouth. Our code can be found at
https://www.github.com/roimehrez/contextualLoss
|
A Hybrid Method for Traffic Flow Forecasting Using Multimodal Deep
Learning | Traffic flow forecasting has been regarded as a key problem of intelligent
transport systems. In this work, we propose a hybrid multimodal deep learning
method for short-term traffic flow forecasting, which can jointly and
adaptively learn the spatial-temporal correlation features and long temporal
interdependence of multi-modality traffic data by an attention auxiliary
multimodal deep learning architecture. According to the highly nonlinear
characteristics of multi-modality traffic data, the base module of our method
consists of one-dimensional Convolutional Neural Networks (1D CNN) and Gated
Recurrent Units (GRU) with the attention mechanism. The former is to capture
the local trend features and the latter is to capture the long temporal
dependencies. Then, we design a hybrid multimodal deep learning framework
(HMDLF) for fusing share representation features of different modality traffic
data by multiple CNN-GRU-Attention modules. The experimental results indicate
that the proposed multimodal deep learning model is capable of dealing with
complex nonlinear urban traffic flow forecasting with satisfying accuracy and
effectiveness.
|
HexaConv | The effectiveness of Convolutional Neural Networks stems in large part from
their ability to exploit the translation invariance that is inherent in many
learning problems. Recently, it was shown that CNNs can exploit other
invariances, such as rotation invariance, by using group convolutions instead
of planar convolutions. However, for reasons of performance and ease of
implementation, it has been necessary to limit the group convolution to
transformations that can be applied to the filters without interpolation. Thus,
for images with square pixels, only integer translations, rotations by
multiples of 90 degrees, and reflections are admissible.
Whereas the square tiling provides a 4-fold rotational symmetry, a hexagonal
tiling of the plane has a 6-fold rotational symmetry. In this paper we show how
one can efficiently implement planar convolution and group convolution over
hexagonal lattices, by re-using existing highly optimized convolution routines.
We find that, due to the reduced anisotropy of hexagonal filters, planar
HexaConv provides better accuracy than planar convolution with square filters,
given a fixed parameter budget. Furthermore, we find that the increased degree
of symmetry of the hexagonal grid increases the effectiveness of group
convolutions, by allowing for more parameter sharing. We show that our method
significantly outperforms conventional CNNs on the AID aerial scene
classification dataset, even outperforming ImageNet pre-trained models.
|
Bidding Machine: Learning to Bid for Directly Optimizing Profits in
Display Advertising | Real-time bidding (RTB) based display advertising has become one of the key
technological advances in computational advertising. RTB enables advertisers to
buy individual ad impressions via an auction in real-time and facilitates the
evaluation and the bidding of individual impressions across multiple
advertisers. In RTB, the advertisers face three main challenges when optimizing
their bidding strategies, namely (i) estimating the utility (e.g., conversions,
clicks) of the ad impression, (ii) forecasting the market value (thus the cost)
of the given ad impression, and (iii) deciding the optimal bid for the given
auction based on the first two. Previous solutions assume the first two are
solved before addressing the bid optimization problem. However, these
challenges are strongly correlated and dealing with any individual problem
independently may not be globally optimal. In this paper, we propose Bidding
Machine, a comprehensive learning to bid framework, which consists of three
optimizers dealing with each challenge above, and as a whole, jointly optimizes
these three parts. We show that such a joint optimization would largely
increase the campaign effectiveness and the profit. From the learning
perspective, we show that the bidding machine can be updated smoothly with both
offline periodical batch or online sequential training schemes. Our extensive
offline empirical study and online A/B testing verify the high effectiveness of
the proposed bidding machine.
|
Relative Pairwise Relationship Constrained Non-negative Matrix
Factorisation | Non-negative Matrix Factorisation (NMF) has been extensively used in machine
learning and data analytics applications. Most existing variations of NMF only
consider how each row/column vector of factorised matrices should be shaped,
and ignore the relationship among pairwise rows or columns. In many cases, such
pairwise relationship enables better factorisation, for example, image
clustering and recommender systems. In this paper, we propose an algorithm
named, Relative Pairwise Relationship constrained Non-negative Matrix
Factorisation (RPR-NMF), which places constraints over relative pairwise
distances amongst features by imposing penalties in a triplet form. Two
distance measures, squared Euclidean distance and Symmetric divergence, are
used, and exponential and hinge loss penalties are adopted for the two measures
respectively. It is well known that the so-called "multiplicative update rules"
result in a much faster convergence than gradient descend for matrix
factorisation. However, applying such update rules to RPR-NMF and also proving
its convergence is not straightforward. Thus, we use reasonable approximations
to relax the complexity brought by the penalties, which are practically
verified. Experiments on both synthetic datasets and real datasets demonstrate
that our algorithms have advantages on gaining close approximation, satisfying
a high proportion of expected constraints, and achieving superior performance
compared with other algorithms.
|
Multi-class Active Learning: A Hybrid Informative and Representative
Criterion Inspired Approach | Labeling each instance in a large dataset is extremely labor- and time-
consuming . One way to alleviate this problem is active learning, which aims to
which discover the most valuable instances for labeling to construct a powerful
classifier. Considering both informativeness and representativeness provides a
promising way to design a practical active learning. However, most existing
active learning methods select instances favoring either informativeness or
representativeness. Meanwhile, many are designed based on the binary class, so
that they may present suboptimal solutions on the datasets with multiple
classes. In this paper, a hybrid informative and representative criterion based
multi-class active learning approach is proposed. We combine the informative
informativeness and representativeness into one formula, which can be solved
under a unified framework. The informativeness is measured by the margin
minimum while the representative information is measured by the maximum mean
discrepancy. By minimizing the upper bound for the true risk, we generalize the
empirical risk minimization principle to the active learning setting.
Simultaneously, our proposed method makes full use of the label information,
and the proposed active learning is designed based on multiple classes. So the
proposed method is not suitable to the binary class but also the multiple
classes. We conduct our experiments on twelve benchmark UCI data sets, and the
experimental results demonstrate that the proposed method performs better than
some state-of-the-art methods.
|
Early Start Intention Detection of Cyclists Using Motion History Images
and a Deep Residual Network | In this article, we present a novel approach to detect starting motions of
cyclists in real world traffic scenarios based on Motion History Images (MHIs).
The method uses a deep Convolutional Neural Network (CNN) with a residual
network architecture (ResNet), which is commonly used in image classification
and detection tasks. By combining MHIs with a ResNet classifier and performing
a frame by frame classification of the MHIs, we are able to detect starting
motions in image sequences. The detection is performed using a wide angle
stereo camera system at an urban intersection. We compare our algorithm to an
existing method to detect movement transitions of pedestrians that uses MHIs in
combination with a Histograms of Oriented Gradients (HOG) like descriptor and a
Support Vector Machine (SVM), which we adapted to cyclists. To train and
evaluate the methods a dataset containing MHIs of 394 cyclist starting motions
was created. The results show that both methods can be used to detect starting
motions of cyclists. Using the SVM approach, we were able to safely detect
starting motions 0.506 s on average after the bicycle starts moving with an
F1-score of 97.7%. The ResNet approach achieved an F1-score of 100% at an
average detection time of 0.144 s. The ResNet approach outperformed the SVM
approach in both robustness against false positive detections and detection
time.
|
MIMO Graph Filters for Convolutional Neural Networks | Superior performance and ease of implementation have fostered the adoption of
Convolutional Neural Networks (CNNs) for a wide array of inference and
reconstruction tasks. CNNs implement three basic blocks: convolution, pooling
and pointwise nonlinearity. Since the two first operations are well-defined
only on regular-structured data such as audio or images, application of CNNs to
contemporary datasets where the information is defined in irregular domains is
challenging. This paper investigates CNNs architectures to operate on signals
whose support can be modeled using a graph. Architectures that replace the
regular convolution with a so-called linear shift-invariant graph filter have
been recently proposed. This paper goes one step further and, under the
framework of multiple-input multiple-output (MIMO) graph filters, imposes
additional structure on the adopted graph filters, to obtain three new (more
parsimonious) architectures. The proposed architectures result in a lower
number of model parameters, reducing the computational complexity, facilitating
the training, and mitigating the risk of overfitting. Simulations show that the
proposed simpler architectures achieve similar performance as more complex
models.
|
Deep Information Networks | We describe a novel classifier with a tree structure, designed using
information theory concepts. This Information Network is made of information
nodes, that compress the input data, and multiplexers, that connect two or more
input nodes to an output node. Each information node is trained, independently
of the others, to minimize a local cost function that minimizes the mutual
information between its input and output with the constraint of keeping a given
mutual information between its output and the target (information bottleneck).
We show that the system is able to provide good results in terms of accuracy,
while it shows many advantages in terms of modularity and reduced complexity.
|
Deep Thermal Imaging: Proximate Material Type Recognition in the Wild
through Deep Learning of Spatial Surface Temperature Patterns | We introduce Deep Thermal Imaging, a new approach for close-range automatic
recognition of materials to enhance the understanding of people and ubiquitous
technologies of their proximal environment. Our approach uses a low-cost mobile
thermal camera integrated into a smartphone to capture thermal textures. A deep
neural network classifies these textures into material types. This approach
works effectively without the need for ambient light sources or direct contact
with materials. Furthermore, the use of a deep learning network removes the
need to handcraft the set of features for different materials. We evaluated the
performance of the system by training it to recognise 32 material types in both
indoor and outdoor environments. Our approach produced recognition accuracies
above 98% in 14,860 images of 15 indoor materials and above 89% in 26,584
images of 17 outdoor materials. We conclude by discussing its potentials for
real-time use in HCI applications and future directions.
|
Dimensionality Reduction for Stationary Time Series via Stochastic
Nonconvex Optimization | Stochastic optimization naturally arises in machine learning. Efficient
algorithms with provable guarantees, however, are still largely missing, when
the objective function is nonconvex and the data points are dependent. This
paper studies this fundamental challenge through a streaming PCA problem for
stationary time series data. Specifically, our goal is to estimate the
principle component of time series data with respect to the covariance matrix
of the stationary distribution. Computationally, we propose a variant of Oja's
algorithm combined with downsampling to control the bias of the stochastic
gradient caused by the data dependency. Theoretically, we quantify the
uncertainty of our proposed stochastic algorithm based on diffusion
approximations. This allows us to prove the asymptotic rate of convergence and
further implies near optimal asymptotic sample complexity. Numerical
experiments are provided to support our analysis.
|
Deep Super Learner: A Deep Ensemble for Classification Problems | Deep learning has become very popular for tasks such as predictive modeling
and pattern recognition in handling big data. Deep learning is a powerful
machine learning method that extracts lower level features and feeds them
forward for the next layer to identify higher level features that improve
performance. However, deep neural networks have drawbacks, which include many
hyper-parameters and infinite architectures, opaqueness into results, and
relatively slower convergence on smaller datasets. While traditional machine
learning algorithms can address these drawbacks, they are not typically capable
of the performance levels achieved by deep neural networks. To improve
performance, ensemble methods are used to combine multiple base learners. Super
learning is an ensemble that finds the optimal combination of diverse learning
algorithms. This paper proposes deep super learning as an approach which
achieves log loss and accuracy results competitive to deep neural networks
while employing traditional machine learning algorithms in a hierarchical
structure. The deep super learner is flexible, adaptable, and easy to train
with good performance across different tasks using identical hyper-parameter
values. Using traditional machine learning requires fewer hyper-parameters,
allows transparency into results, and has relatively fast convergence on
smaller datasets. Experimental results show that the deep super learner has
superior performance compared to the individual base learners, single-layer
ensembles, and in some cases deep neural networks. Performance of the deep
super learner may further be improved with task-specific tuning.
|
Learning Memory Access Patterns | The explosion in workload complexity and the recent slow-down in Moore's law
scaling call for new approaches towards efficient computing. Researchers are
now beginning to use recent advances in machine learning in software
optimizations, augmenting or replacing traditional heuristics and data
structures. However, the space of machine learning for computer hardware
architecture is only lightly explored. In this paper, we demonstrate the
potential of deep learning to address the von Neumann bottleneck of memory
performance. We focus on the critical problem of learning memory access
patterns, with the goal of constructing accurate and efficient memory
prefetchers. We relate contemporary prefetching strategies to n-gram models in
natural language processing, and show how recurrent neural networks can serve
as a drop-in replacement. On a suite of challenging benchmark datasets, we find
that neural networks consistently demonstrate superior performance in terms of
precision and recall. This work represents the first step towards practical
neural-network based prefetching, and opens a wide range of exciting directions
for machine learning in computer architecture research.
|
Smoothed Action Value Functions for Learning Gaussian Policies | State-action value functions (i.e., Q-values) are ubiquitous in reinforcement
learning (RL), giving rise to popular algorithms such as SARSA and Q-learning.
We propose a new notion of action value defined by a Gaussian smoothed version
of the expected Q-value. We show that such smoothed Q-values still satisfy a
Bellman equation, making them learnable from experience sampled from an
environment. Moreover, the gradients of expected reward with respect to the
mean and covariance of a parameterized Gaussian policy can be recovered from
the gradient and Hessian of the smoothed Q-value function. Based on these
relationships, we develop new algorithms for training a Gaussian policy
directly from a learned smoothed Q-value approximator. The approach is
additionally amenable to proximal optimization by augmenting the objective with
a penalty on KL-divergence from a previous policy. We find that the ability to
learn both a mean and covariance during training leads to significantly
improved results on standard continuous control benchmarks.
|
Learning SMaLL Predictors | We present a new machine learning technique for training small
resource-constrained predictors. Our algorithm, the Sparse Multiprototype
Linear Learner (SMaLL), is inspired by the classic machine learning problem of
learning $k$-DNF Boolean formulae. We present a formal derivation of our
algorithm and demonstrate the benefits of our approach with a detailed
empirical study.
|
Arbitrary Discrete Sequence Anomaly Detection with Zero Boundary LSTM | We propose a simple mathematical definition and new neural architecture for
finding anomalies within discrete sequence datasets. Our model comprises of a
modified LSTM autoencoder and an array of One-Class SVMs. The LSTM takes in
elements from a sequence and creates context vectors that are used to predict
the probability distribution of the following element. These context vectors
are then used to train an array of One-Class SVMs. These SVMs are used to
determine an outlier boundary in context space.We show that our method is
consistently more stable and also outperforms standard LSTM and sliding window
anomaly detection systems on two generated datasets.
|
Visualizing Convolutional Neural Network Protein-Ligand Scoring | Protein-ligand scoring is an important step in a structure-based drug design
pipeline. Selecting a correct binding pose and predicting the binding affinity
of a protein-ligand complex enables effective virtual screening. Machine
learning techniques can make use of the increasing amounts of structural data
that are becoming publicly available. Convolutional neural network (CNN)
scoring functions in particular have shown promise in pose selection and
affinity prediction for protein-ligand complexes. Neural networks are known for
being difficult to interpret. Understanding the decisions of a particular
network can help tune parameters and training data to maximize performance.
Visualization of neural networks helps decompose complex scoring functions into
pictures that are more easily parsed by humans. Here we present three methods
for visualizing how individual protein-ligand complexes are interpreted by 3D
convolutional neural networks. We also present a visualization of the
convolutional filters and their weights. We describe how the intuition provided
by these visualizations aids in network design.
|
Natural Language to Structured Query Generation via Meta-Learning | In conventional supervised training, a model is trained to fit all the
training examples. However, having a monolithic model may not always be the
best strategy, as examples could vary widely. In this work, we explore a
different learning protocol that treats each example as a unique pseudo-task,
by reducing the original learning problem to a few-shot meta-learning scenario
with the help of a domain-dependent relevance function. When evaluated on the
WikiSQL dataset, our approach leads to faster convergence and achieves
1.1%-5.4% absolute accuracy gains over the non-meta-learning counterparts.
|
Masked Conditional Neural Networks for Audio Classification | We present the ConditionaL Neural Network (CLNN) and the Masked ConditionaL
Neural Network (MCLNN) designed for temporal signal recognition. The CLNN takes
into consideration the temporal nature of the sound signal and the MCLNN
extends upon the CLNN through a binary mask to preserve the spatial locality of
the features and allows an automated exploration of the features combination
analogous to hand-crafting the most relevant features for the recognition task.
MCLNN has achieved competitive recognition accuracies on the GTZAN and the
ISMIR2004 music datasets that surpass several state-of-the-art neural network
based architectures and hand-crafted methods applied on both datasets.
|
A Reductions Approach to Fair Classification | We present a systematic approach for achieving fairness in a binary
classification setting. While we focus on two well-known quantitative
definitions of fairness, our approach encompasses many other previously studied
definitions as special cases. The key idea is to reduce fair classification to
a sequence of cost-sensitive classification problems, whose solutions yield a
randomized classifier with the lowest (empirical) error subject to the desired
constraints. We introduce two reductions that work for any representation of
the cost-sensitive classifier and compare favorably to prior baselines on a
variety of data sets, while overcoming several of their disadvantages.
|
Robust Multiple Kernel k-means Clustering using Min-Max Optimization | Multiple kernel learning is a type of multiview learning that combines
different data modalities by capturing view-specific patterns using kernels.
Although supervised multiple kernel learning has been extensively studied,
until recently, only a few unsupervised approaches have been proposed. In the
meanwhile, adversarial learning has recently received much attention. Many
works have been proposed to defend against adversarial examples. However,
little is known about the effect of adversarial perturbation in the context of
multiview learning, and even less in the unsupervised case. In this study, we
show that adversarial features added to a view can make the existing approaches
with the min-max formulation in multiple kernel clustering yield unfavorable
clusters. To address this problem and inspired by recent works in adversarial
learning, we propose a multiple kernel clustering method with the min-max
framework that aims to be robust to such adversarial perturbation. We evaluate
the robustness of our method on simulation data under different types of
adversarial perturbations and show that it outperforms several compared
existing methods. In the real data analysis, We demonstrate the utility of our
method on a real-world problem.
|
Discontinuity-Sensitive Optimal Control Learning by Mixture of Experts | This paper proposes a discontinuity-sensitive approach to learn the solutions
of parametric optimal control problems with high accuracy. Many tasks, ranging
from model predictive control to reinforcement learning, may be solved by
learning optimal solutions as a function of problem parameters. However,
nonconvexity, discrete homotopy classes, and control switching cause
discontinuity in the parameter-solution mapping, thus making learning difficult
for traditional continuous function approximators. A mixture of experts (MoE)
model composed of a classifier and several regressors is proposed to address
such an issue. The optimal trajectories of different parameters are clustered
such that in each cluster the trajectories are continuous function of problem
parameters. Numerical examples on benchmark problems show that training the
classifier and regressors individually outperforms joint training of MoE. With
suitably chosen clusters, this approach not only achieves lower prediction
error with less training data and fewer model parameters, but also leads to
dramatic improvements in the reliability of trajectory tracking compared to
traditional universal function approximation models (e.g., neural networks).
|
Exponential Discriminative Metric Embedding in Deep Learning | With the remarkable success achieved by the Convolutional Neural Networks
(CNNs) in object recognition recently, deep learning is being widely used in
the computer vision community. Deep Metric Learning (DML), integrating deep
learning with conventional metric learning, has set new records in many fields,
especially in classification task. In this paper, we propose a replicable DML
method, called Include and Exclude (IE) loss, to force the distance between a
sample and its designated class center away from the mean distance of this
sample to other class centers with a large margin in the exponential feature
projection space. With the supervision of IE loss, we can train CNNs to enhance
the intra-class compactness and inter-class separability, leading to great
improvements on several public datasets ranging from object recognition to face
verification. We conduct a comparative study of our algorithm with several
typical DML methods on three kinds of networks with different capacity.
Extensive experiments on three object recognition datasets and two face
recognition datasets demonstrate that IE loss is always superior to other
mainstream DML methods and approach the state-of-the-art results.
|
Sequential Maximum Margin Classifiers for Partially Labeled Data | In many real-world applications, data is not collected as one batch, but
sequentially over time, and often it is not possible or desirable to wait until
the data is completely gathered before analyzing it. Thus, we propose a
framework to sequentially update a maximum margin classifier by taking
advantage of the Maximum Entropy Discrimination principle. Our maximum margin
classifier allows for a kernel representation to represent large numbers of
features and can also be regularized with respect to a smooth sub-manifold,
allowing it to incorporate unlabeled observations. We compare the performance
of our classifier to its non-sequential equivalents in both simulated and real
datasets.
|
Visual Explanations From Deep 3D Convolutional Neural Networks for
Alzheimer's Disease Classification | We develop three efficient approaches for generating visual explanations from
3D convolutional neural networks (3D-CNNs) for Alzheimer's disease
classification. One approach conducts sensitivity analysis on hierarchical 3D
image segmentation, and the other two visualize network activations on a
spatial map. Visual checks and a quantitative localization benchmark indicate
that all approaches identify important brain parts for Alzheimer's disease
diagnosis. Comparative analysis show that the sensitivity analysis based
approach has difficulty handling loosely distributed cerebral cortex, and
approaches based on visualization of activations are constrained by the
resolution of the convolutional layer. The complementarity of these methods
improves the understanding of 3D-CNNs in Alzheimer's disease classification
from different perspectives.
|
Extracting Domain Invariant Features by Unsupervised Learning for Robust
Automatic Speech Recognition | The performance of automatic speech recognition (ASR) systems can be
significantly compromised by previously unseen conditions, which is typically
due to a mismatch between training and testing distributions. In this paper, we
address robustness by studying domain invariant features, such that domain
information becomes transparent to ASR systems, resolving the mismatch problem.
Specifically, we investigate a recent model, called the Factorized Hierarchical
Variational Autoencoder (FHVAE). FHVAEs learn to factorize sequence-level and
segment-level attributes into different latent variables without supervision.
We argue that the set of latent variables that contain segment-level
information is our desired domain invariant feature for ASR. Experiments are
conducted on Aurora-4 and CHiME-4, which demonstrate 41% and 27% absolute word
error rate reductions respectively on mismatched domains.
|
Graph Learning from Filtered Signals: Graph System and Diffusion Kernel
Identification | This paper introduces a novel graph signal processing framework for building
graph-based models from classes of filtered signals. In our framework,
graph-based modeling is formulated as a graph system identification problem,
where the goal is to learn a weighted graph (a graph Laplacian matrix) and a
graph-based filter (a function of graph Laplacian matrices). In order to solve
the proposed problem, an algorithm is developed to jointly identify a graph and
a graph-based filter (GBF) from multiple signal/data observations. Our
algorithm is valid under the assumption that GBFs are one-to-one functions. The
proposed approach can be applied to learn diffusion (heat) kernels, which are
popular in various fields for modeling diffusion processes. In addition, for
specific choices of graph-based filters, the proposed problem reduces to a
graph Laplacian estimation problem. Our experimental results demonstrate that
the proposed algorithm outperforms the current state-of-the-art methods. We
also implement our framework on a real climate dataset for modeling of
temperature signals.
|
Revisiting differentially private linear regression: optimal and
adaptive prediction & estimation in unbounded domain | We revisit the problem of linear regression under a differential privacy
constraint. By consolidating existing pieces in the literature, we clarify the
correct dependence of the feature, label and coefficient domains in the
optimization error and estimation error, hence revealing the delicate price of
differential privacy in statistical estimation and statistical learning.
Moreover, we propose simple modifications of two existing DP algorithms: (a)
posterior sampling, (b) sufficient statistics perturbation, and show that they
can be upgraded into **adaptive** algorithms that are able to exploit
data-dependent quantities and behave nearly optimally **for every instance**.
Extensive experiments are conducted on both simulated data and real data, which
conclude that both AdaOPS and AdaSSP outperform the existing techniques on
nearly all 36 data sets that we test on.
|
Gaussian Process Latent Variable Alignment Learning | We present a model that can automatically learn alignments between
high-dimensional data in an unsupervised manner. Our proposed method casts
alignment learning in a framework where both alignment and data are modelled
simultaneously. Further, we automatically infer groupings of different types of
sequences within the same dataset. We derive a probabilistic model built on
non-parametric priors that allows for flexible warps while at the same time
providing means to specify interpretable constraints. We demonstrate the
efficacy of our approach with superior quantitative performance to the
state-of-the-art approaches and provide examples to illustrate the versatility
of our model in automatic inference of sequence groupings, absent from previous
approaches, as well as easy specification of high level priors for different
modalities of data.
|
Sketching for Principal Component Regression | Principal component regression (PCR) is a useful method for regularizing
linear regression. Although conceptually simple, straightforward
implementations of PCR have high computational costs and so are inappropriate
when learning with large scale data. In this paper, we propose efficient
algorithms for computing approximate PCR solutions that are, on one hand, high
quality approximations to the true PCR solutions (when viewed as minimizer of a
constrained optimization problem), and on the other hand entertain rigorous
risk bounds (when viewed as statistical estimators). In particular, we propose
an input sparsity time algorithms for approximate PCR. We also consider
computing an approximate PCR in the streaming model, and kernel PCR. Empirical
results demonstrate the excellent performance of our proposed methods.
|
A Neural Network Approach to Missing Marker Reconstruction in Human
Motion Capture | Optical motion capture systems have become a widely used technology in
various fields, such as augmented reality, robotics, movie production, etc.
Such systems use a large number of cameras to triangulate the position of
optical markers.The marker positions are estimated with high accuracy. However,
especially when tracking articulated bodies, a fraction of the markers in each
timestep is missing from the reconstruction. In this paper, we propose to use a
neural network approach to learn how human motion is temporally and spatially
correlated, and reconstruct missing markers positions through this model. We
experiment with two different models, one LSTM-based and one time-window-based.
Both methods produce state-of-the-art results, while working online, as opposed
to most of the alternative methods, which require the complete sequence to be
known. The implementation is publicly available at
https://github.com/Svito-zar/NN-for-Missing-Marker-Reconstruction .
|
Transfer Learning with Neural AutoML | We reduce the computational cost of Neural AutoML with transfer learning.
AutoML relieves human effort by automating the design of ML algorithms. Neural
AutoML has become popular for the design of deep learning architectures,
however, this method has a high computation cost. To address this we propose
Transfer Neural AutoML that uses knowledge from prior tasks to speed up network
design. We extend RL-based architecture search methods to support parallel
training on multiple tasks and then transfer the search strategy to new tasks.
On language and image classification tasks, Transfer Neural AutoML reduces
convergence time over single-task training by over an order of magnitude on
many tasks.
|
Fast Dawid-Skene: A Fast Vote Aggregation Scheme for Sentiment
Classification | Many real world problems can now be effectively solved using supervised
machine learning. A major roadblock is often the lack of an adequate quantity
of labeled data for training. A possible solution is to assign the task of
labeling data to a crowd, and then infer the true label using aggregation
methods. A well-known approach for aggregation is the Dawid-Skene (DS)
algorithm, which is based on the principle of Expectation-Maximization (EM). We
propose a new simple, yet effective, EM-based algorithm, which can be
interpreted as a `hard' version of DS, that allows much faster convergence
while maintaining similar accuracy in aggregation. We show the use of this
algorithm as a quick and effective technique for online, real-time sentiment
annotation. We also prove that our algorithm converges to the estimated labels
at a linear rate. Our experiments on standard datasets show a significant
speedup in time taken for aggregation - upto $\sim$8x over Dawid-Skene and
$\sim$6x over other fast EM methods, at competitive accuracy performance. The
code for the implementation of the algorithms can be found at
https://github.com/GoodDeeds/Fast-Dawid-Skene
|
A bag-to-class divergence approach to multiple-instance learning | In multi-instance (MI) learning, each object (bag) consists of multiple
feature vectors (instances), and is most commonly regarded as a set of points
in a multidimensional space. A different viewpoint is that the instances are
realisations of random vectors with corresponding probability distribution, and
that a bag is the distribution, not the realisations. In MI classification,
each bag in the training set has a class label, but the instances are
unlabelled. By introducing the probability distribution space to bag-level
classification problems, dissimilarities between probability distributions
(divergences) can be applied. The bag-to-bag Kullback-Leibler information is
asymptotically the best classifier, but the typical sparseness of MI training
sets is an obstacle. We introduce bag-to-class divergence to MI learning,
emphasising the hierarchical nature of the random vectors that makes bags from
the same class different. We propose two properties for bag-to-class
divergences, and an additional property for sparse training sets.
|
Accelerated Methods for Deep Reinforcement Learning | Deep reinforcement learning (RL) has achieved many recent successes, yet
experiment turn-around time remains a key bottleneck in research and in
practice. We investigate how to optimize existing deep RL algorithms for modern
computers, specifically for a combination of CPUs and GPUs. We confirm that
both policy gradient and Q-value learning algorithms can be adapted to learn
using many parallel simulator instances. We further find it possible to train
using batch sizes considerably larger than are standard, without negatively
affecting sample complexity or final performance. We leverage these facts to
build a unified framework for parallelization that dramatically hastens
experiments in both classes of algorithm. All neural network computations use
GPUs, accelerating both data collection and training. Our results include using
an entire DGX-1 to learn successful strategies in Atari games in mere minutes,
using both synchronous and asynchronous algorithms.
|
Sever: A Robust Meta-Algorithm for Stochastic Optimization | In high dimensions, most machine learning methods are brittle to even a small
fraction of structured outliers. To address this, we introduce a new
meta-algorithm that can take in a base learner such as least squares or
stochastic gradient descent, and harden the learner to be resistant to
outliers. Our method, Sever, possesses strong theoretical guarantees yet is
also highly scalable -- beyond running the base learner itself, it only
requires computing the top singular vector of a certain $n \times d$ matrix. We
apply Sever on a drug design dataset and a spam classification dataset, and
find that in both cases it has substantially greater robustness than several
baselines. On the spam dataset, with $1\%$ corruptions, we achieved $7.4\%$
test error, compared to $13.4\%-20.5\%$ for the baselines, and $3\%$ error on
the uncorrupted dataset. Similarly, on the drug design dataset, with $10\%$
corruptions, we achieved $1.42$ mean-squared error test error, compared to
$1.51$-$2.33$ for the baselines, and $1.23$ error on the uncorrupted dataset.
|
Satisficing in Time-Sensitive Bandit Learning | Much of the recent literature on bandit learning focuses on algorithms that
aim to converge on an optimal action. One shortcoming is that this orientation
does not account for time sensitivity, which can play a crucial role when
learning an optimal action requires much more information than near-optimal
ones. Indeed, popular approaches such as upper-confidence-bound methods and
Thompson sampling can fare poorly in such situations. We consider instead
learning a satisficing action, which is near-optimal while requiring less
information, and propose satisficing Thompson sampling, an algorithm that
serves this purpose. We establish a general bound on expected discounted regret
and study the application of satisficing Thompson sampling to linear and
infinite-armed bandits, demonstrating arbitrarily large benefits over Thompson
sampling. We also discuss the relation between the notion of satisficing and
the theory of rate distortion, which offers guidance on the selection of
satisficing actions.
|
WNGrad: Learn the Learning Rate in Gradient Descent | Adjusting the learning rate schedule in stochastic gradient methods is an
important unresolved problem which requires tuning in practice. If certain
parameters of the loss function such as smoothness or strong convexity
constants are known, theoretical learning rate schedules can be applied.
However, in practice, such parameters are not known, and the loss function of
interest is not convex in any case. The recently proposed batch normalization
reparametrization is widely adopted in most neural network architectures today
because, among other advantages, it is robust to the choice of Lipschitz
constant of the gradient in loss function, allowing one to set a large learning
rate without worry. Inspired by batch normalization, we propose a general
nonlinear update rule for the learning rate in batch and stochastic gradient
descent so that the learning rate can be initialized at a high value, and is
subsequently decreased according to gradient observations along the way. The
proposed method is shown to achieve robustness to the relationship between the
learning rate and the Lipschitz constant, and near-optimal convergence rates in
both the batch and stochastic settings ($O(1/T)$ for smooth loss in the batch
setting, and $O(1/\sqrt{T})$ for convex loss in the stochastic setting). We
also show through numerical evidence that such robustness of the proposed
method extends to highly nonconvex and possibly non-smooth loss function in
deep learning problems.Our analysis establishes some first theoretical
understanding into the observed robustness for batch normalization and weight
normalization.
|
Deep Models of Interactions Across Sets | We use deep learning to model interactions across two or more sets of
objects, such as user-movie ratings, protein-drug bindings, or ternary
user-item-tag interactions. The canonical representation of such interactions
is a matrix (or a higher-dimensional tensor) with an exchangeability property:
the encoding's meaning is not changed by permuting rows or columns. We argue
that models should hence be Permutation Equivariant (PE): constrained to make
the same predictions across such permutations. We present a parameter-sharing
scheme and prove that it could not be made any more expressive without
violating PE. This scheme yields three benefits. First, we demonstrate
state-of-the-art performance on multiple matrix completion benchmarks. Second,
our models require a number of parameters independent of the numbers of
objects, and thus scale well to large datasets. Third, models can be queried
about new objects that were not available at training time, but for which
interactions have since been observed. In experiments, our models achieved
surprisingly good generalization performance on this matrix extrapolation task,
both within domains (e.g., new users and new movies drawn from the same
distribution used for training) and even across domains (e.g., predicting music
ratings after training on movies).
|
An efficient framework for learning sentence representations | In this work we propose a simple and efficient framework for learning
sentence representations from unlabelled data. Drawing inspiration from the
distributional hypothesis and recent work on learning sentence representations,
we reformulate the problem of predicting the context in which a sentence
appears as a classification problem. Given a sentence and its context, a
classifier distinguishes context sentences from other contrastive sentences
based on their vector representations. This allows us to efficiently learn
different types of encoding functions, and we show that the model learns
high-quality sentence representations. We demonstrate that our sentence
representations outperform state-of-the-art unsupervised and supervised
representation learning methods on several downstream NLP tasks that involve
understanding sentence semantics while achieving an order of magnitude speedup
in training time.
|
Fast Convergence for Stochastic and Distributed Gradient Descent in the
Interpolation Limit | Modern supervised learning techniques, particularly those using deep nets,
involve fitting high dimensional labelled data sets with functions containing
very large numbers of parameters. Much of this work is empirical. Interesting
phenomena have been observed that require theoretical explanations; however the
non-convexity of the loss functions complicates the analysis. Recently it has
been proposed that the success of these techniques rests partly in the
effectiveness of the simple stochastic gradient descent algorithm in the so
called interpolation limit in which all labels are fit perfectly. This analysis
is made possible since the SGD algorithm reduces to a stochastic linear system
near the interpolating minimum of the loss function. Here we exploit this
insight by presenting and analyzing a new distributed algorithm for gradient
descent, also in the interpolating limit. The distributed SGD algorithm
presented in the paper corresponds to gradient descent applied to a simple
penalized distributed loss function, $L({\bf w}_1,...,{\bf w}_n) = \Sigma_i
l_i({\bf w}_i) + \mu \sum_{<i,j>}|{\bf w}_i-{\bf w}_j|^2$. Here each node holds
only one sample, and its own parameter vector. The notation $<i,j>$ denotes
edges of a connected graph defining the links between nodes. It is shown that
this distributed algorithm converges linearly (ie the error reduces
exponentially with iteration number), with a rate
$1-\frac{\eta}{n}\lambda_{min}(H)<R<1$ where $\lambda_{min}(H)$ is the smallest
nonzero eigenvalue of the sample covariance or the Hessian H. In contrast with
previous usage of similar penalty functions to enforce consensus between nodes,
in the interpolating limit it is not required to take the penalty parameter to
infinity for consensus to occur. The analysis further reinforces the utility of
the interpolation limit in the theoretical treatment of modern machine learning
algorithms.
|
The Advantage of Doubling: A Deep Reinforcement Learning Approach to
Studying the Double Team in the NBA | During the 2017 NBA playoffs, Celtics coach Brad Stevens was faced with a
difficult decision when defending against the Cavaliers: "Do you double and
risk giving up easy shots, or stay at home and do the best you can?" It's a
tough call, but finding a good defensive strategy that effectively incorporates
doubling can make all the difference in the NBA. In this paper, we analyze
double teaming in the NBA, quantifying the trade-off between risk and reward.
Using player trajectory data pertaining to over 643,000 possessions, we
identified when the ball handler was double teamed. Given these data and the
corresponding outcome (i.e., was the defense successful), we used deep
reinforcement learning to estimate the quality of the defensive actions. We
present qualitative and quantitative results summarizing our learned defensive
strategy for defending. We show that our policy value estimates are predictive
of points per possession and win percentage. Overall, the proposed framework
represents a step toward a more comprehensive understanding of defensive
strategies in the NBA.
|
Some Approximation Bounds for Deep Networks | In this paper we introduce new bounds on the approximation of functions in
deep networks and in doing so introduce some new deep network architectures for
function approximation. These results give some theoretical insight into the
success of autoencoders and ResNets.
|
A Multi-Objective Deep Reinforcement Learning Framework | This paper introduces a new scalable multi-objective deep reinforcement
learning (MODRL) framework based on deep Q-networks. We develop a
high-performance MODRL framework that supports both single-policy and
multi-policy strategies, as well as both linear and non-linear approaches to
action selection. The experimental results on two benchmark problems
(two-objective deep sea treasure environment and three-objective Mountain Car
problem) indicate that the proposed framework is able to find the
Pareto-optimal solutions effectively. The proposed framework is generic and
highly modularized, which allows the integration of different deep
reinforcement learning algorithms in different complex problem domains. This
therefore overcomes many disadvantages involved with standard multi-objective
reinforcement learning methods in the current literature. The proposed
framework acts as a testbed platform that accelerates the development of MODRL
for solving increasingly complicated multi-objective problems.
|
Disentangled Sequential Autoencoder | We present a VAE architecture for encoding and generating high dimensional
sequential data, such as video or audio. Our deep generative model learns a
latent representation of the data which is split into a static and dynamic
part, allowing us to approximately disentangle latent time-dependent features
(dynamics) from features which are preserved over time (content). This
architecture gives us partial control over generating content and dynamics by
conditioning on either one of these sets of features. In our experiments on
artificially generated cartoon video clips and voice recordings, we show that
we can convert the content of a given sequence into another one by such content
swapping. For audio, this allows us to convert a male speaker into a female
speaker and vice versa, while for video we can separately manipulate shapes and
dynamics. Furthermore, we give empirical evidence for the hypothesis that
stochastic RNNs as latent state models are more efficient at compressing and
generating long sequences than deterministic ones, which may be relevant for
applications in video compression.
|
On First-Order Meta-Learning Algorithms | This paper considers meta-learning problems, where there is a distribution of
tasks, and we would like to obtain an agent that performs well (i.e., learns
quickly) when presented with a previously unseen task sampled from this
distribution. We analyze a family of algorithms for learning a parameter
initialization that can be fine-tuned quickly on a new task, using only
first-order derivatives for the meta-learning updates. This family includes and
generalizes first-order MAML, an approximation to MAML obtained by ignoring
second-order derivatives. It also includes Reptile, a new algorithm that we
introduce here, which works by repeatedly sampling a task, training on it, and
moving the initialization towards the trained weights on that task. We expand
on the results from Finn et al. showing that first-order meta-learning
algorithms perform well on some well-established benchmarks for few-shot
classification, and we provide theoretical analysis aimed at understanding why
these algorithms work.
|
Cross-domain Recommendation via Deep Domain Adaptation | The behavior of users in certain services could be a clue that can be used to
infer their preferences and may be used to make recommendations for other
services they have never used. However, the cross-domain relationships between
items and user consumption patterns are not simple, especially when there are
few or no common users and items across domains. To address this problem, we
propose a content-based cross-domain recommendation method for cold-start users
that does not require user- and item- overlap. We formulate recommendation as
extreme multi-class classification where labels (items) corresponding to the
users are predicted. With this formulation, the problem is reduced to a domain
adaptation setting, in which a classifier trained in the source domain is
adapted to the target domain. For this, we construct a neural network that
combines an architecture for domain adaptation, Domain Separation Network, with
a denoising autoencoder for item representation. We assess the performance of
our approach in experiments on a pair of data sets collected from movie and
news services of Yahoo! JAPAN and show that our approach outperforms several
baseline methods including a cross-domain collaborative filtering method.
|
Design of a nickel-base superalloy using a neural network | A new computational tool has been developed to model, discover, and optimize
new alloys that simultaneously satisfy up to eleven physical criteria. An
artificial neural network is trained from pre-existing materials data that
enables the prediction of individual material properties both as a function of
composition and heat treatment routine, which allows it to optimize the
material properties to search for the material with properties most likely to
exceed a target criteria. We design a new polycrystalline nickel-base
superalloy with the optimal combination of cost, density, gamma' phase content
and solvus, phase stability, fatigue life, yield stress, ultimate tensile
strength, stress rupture, oxidation resistance, and tensile elongation.
Experimental data demonstrates that the proposed alloy fulfills the
computational predictions, possessing multiple physical properties,
particularly oxidation resistance and yield stress, that exceed existing
commercially available alloys.
|
Physical Layer Communications System Design Over-the-Air Using
Adversarial Networks | This paper presents a novel method for synthesizing new physical layer
modulation and coding schemes for communications systems using a learning-based
approach which does not require an analytic model of the impairments in the
channel. It extends prior work published on the channel autoencoder to consider
the case where the channel response is not known or can not be easily modeled
in a closed form analytic expression. By adopting an adversarial approach for
channel response approximation and information encoding, we can jointly learn a
good solution to both tasks over a wide range of channel environments. We
describe the operation of the proposed adversarial system, share results for
its training and validation over-the-air, and discuss implications and future
work in the area.
|
Generating Artificial Data for Private Deep Learning | In this paper, we propose generating artificial data that retain statistical
properties of real data as the means of providing privacy with respect to the
original dataset. We use generative adversarial network to draw
privacy-preserving artificial data samples and derive an empirical method to
assess the risk of information disclosure in a differential-privacy-like way.
Our experiments show that we are able to generate artificial data of high
quality and successfully train and validate machine learning models on this
data while limiting potential privacy loss.
|
Learning Rules-First Classifiers | Complex classifiers may exhibit "embarassing" failures in cases where humans
can easily provide a justified classification. Avoiding such failures is
obviously of key importance. In this work, we focus on one such setting, where
a label is perfectly predictable if the input contains certain features, or
rules, and otherwise it is predictable by a linear classifier. We define a
hypothesis class that captures this notion and determine its sample complexity.
We also give evidence that efficient algorithms cannot achieve this sample
complexity. We then derive a simple and efficient algorithm and show that its
sample complexity is close to optimal, among efficient algorithms. Experiments
on synthetic and sentiment analysis data demonstrate the efficacy of the
method, both in terms of accuracy and interpretability.
|
Drug Recommendation toward Safe Polypharmacy | Adverse drug reactions (ADRs) induced from high-order drug-drug interactions
(DDIs) due to polypharmacy represent a significant public health problem. In
this paper, we formally formulate the to-avoid and safe (with respect to ADRs)
drug recommendation problems when multiple drugs have been taken
simultaneously. We develop a joint model with a recommendation component and an
ADR label prediction component to recommend for a prescription a set of
to-avoid drugs that will induce ADRs if taken together with the prescription.
We also develop real drug-drug interaction datasets and corresponding
evaluation protocols. Our experimental results on real datasets demonstrate the
strong performance of the joint model compared to other baseline methods.
|
A Bayesian and Machine Learning approach to estimating Influence Model
parameters for IM-RO | The rise of Online Social Networks (OSNs) has caused an insurmountable amount
of interest from advertisers and researchers seeking to monopolize on its
features. Researchers aim to develop strategies for determining how information
is propagated among users within an OSN that is captured by diffusion or
influence models. We consider the influence models for the IM-RO problem, a
novel formulation to the Influence Maximization (IM) problem based on
implementing Stochastic Dynamic Programming (SDP). In contrast to existing
approaches involving influence spread and the theory of submodular functions,
the SDP method focuses on optimizing clicks and ultimately revenue to
advertisers in OSNs. Existing approaches to influence maximization have been
actively researched over the past decade, with applications to multiple fields,
however, our approach is a more practical variant to the original IM problem.
In this paper, we provide an analysis on the influence models of the IM-RO
problem by conducting experiments on synthetic and real-world datasets. We
propose a Bayesian and Machine Learning approach for estimating the parameters
of the influence models for the (Influence Maximization- Revenue Optimization)
IM-RO problem. We present a Bayesian hierarchical model and implement the
well-known Naive Bayes classifier (NBC), Decision Trees classifier (DTC) and
Random Forest classifier (RFC) on three real-world datasets. Compared to
previous approaches to estimating influence model parameters, our strategy has
the great advantage of being directly implementable in standard software
packages such as WinBUGS/OpenBUGS/JAGS and Apache Spark. We demonstrate the
efficiency and usability of our methods in terms of spreading information and
generating revenue for advertisers in the context of OSNs.
|
Improving Optimization for Models With Continuous Symmetry Breaking | Many loss functions in representation learning are invariant under a
continuous symmetry transformation. For example, the loss function of word
embeddings (Mikolov et al., 2013) remains unchanged if we simultaneously rotate
all word and context embedding vectors. We show that representation learning
models for time series possess an approximate continuous symmetry that leads to
slow convergence of gradient descent. We propose a new optimization algorithm
that speeds up convergence using ideas from gauge theory in physics. Our
algorithm leads to orders of magnitude faster convergence and to more
interpretable representations, as we show for dynamic extensions of matrix
factorization and word embedding models. We further present an example
application of our proposed algorithm that translates modern words into their
historic equivalents.
|
Fairness Through Computationally-Bounded Awareness | We study the problem of fair classification within the versatile framework of
Dwork et al. [ITCS '12], which assumes the existence of a metric that measures
similarity between pairs of individuals. Unlike earlier work, we do not assume
that the entire metric is known to the learning algorithm; instead, the learner
can query this arbitrary metric a bounded number of times. We propose a new
notion of fairness called metric multifairness and show how to achieve this
notion in our setting. Metric multifairness is parameterized by a similarity
metric $d$ on pairs of individuals to classify and a rich collection ${\cal C}$
of (possibly overlapping) "comparison sets" over pairs of individuals. At a
high level, metric multifairness guarantees that similar subpopulations are
treated similarly, as long as these subpopulations are identified within the
class ${\cal C}$.
|
Efficient Algorithms for Outlier-Robust Regression | We give the first polynomial-time algorithm for performing linear or
polynomial regression resilient to adversarial corruptions in both examples and
labels.
Given a sufficiently large (polynomial-size) training set drawn i.i.d. from
distribution D and subsequently corrupted on some fraction of points, our
algorithm outputs a linear function whose squared error is close to the squared
error of the best-fitting linear function with respect to D, assuming that the
marginal distribution of D over the input space is \emph{certifiably
hypercontractive}. This natural property is satisfied by many well-studied
distributions such as Gaussian, strongly log-concave distributions and, uniform
distribution on the hypercube among others. We also give a simple statistical
lower bound showing that some distributional assumption is necessary to succeed
in this setting.
These results are the first of their kind and were not known to be even
information-theoretically possible prior to our work.
Our approach is based on the sum-of-squares (SoS) method and is inspired by
the recent applications of the method for parameter recovery problems in
unsupervised learning. Our algorithm can be seen as a natural convex relaxation
of the following conceptually simple non-convex optimization problem: find a
linear function and a large subset of the input corrupted sample such that the
least squares loss of the function over the subset is minimized over all
possible large subsets.
|
Probably Approximately Metric-Fair Learning | The seminal work of Dwork {\em et al.} [ITCS 2012] introduced a metric-based
notion of individual fairness. Given a task-specific similarity metric, their
notion required that every pair of similar individuals should be treated
similarly. In the context of machine learning, however, individual fairness
does not generalize from a training set to the underlying population. We show
that this can lead to computational intractability even for simple
fair-learning tasks.
With this motivation in mind, we introduce and study a relaxed notion of {\em
approximate metric-fairness}: for a random pair of individuals sampled from the
population, with all but a small probability of error, if they are similar then
they should be treated similarly. We formalize the goal of achieving
approximate metric-fairness simultaneously with best-possible accuracy as
Probably Approximately Correct and Fair (PACF) Learning. We show that
approximate metric-fairness {\em does} generalize, and leverage these
generalization guarantees to construct polynomial-time PACF learning algorithms
for the classes of linear and logistic predictors.
|
GONet: A Semi-Supervised Deep Learning Approach For Traversability
Estimation | We present semi-supervised deep learning approaches for traversability
estimation from fisheye images. Our method, GONet, and the proposed extensions
leverage Generative Adversarial Networks (GANs) to effectively predict whether
the area seen in the input image(s) is safe for a robot to traverse. These
methods are trained with many positive images of traversable places, but just a
small set of negative images depicting blocked and unsafe areas. This makes the
proposed methods practical. Positive examples can be collected easily by simply
operating a robot through traversable spaces, while obtaining negative examples
is time consuming, costly, and potentially dangerous. Through extensive
experiments and several demonstrations, we show that the proposed
traversability estimation approaches are robust and can generalize to unseen
scenarios. Further, we demonstrate that our methods are memory efficient and
fast, allowing for real-time operation on a mobile robot with single or stereo
fisheye cameras. As part of our contributions, we open-source two new datasets
for traversability estimation. These datasets are composed of approximately 24h
of videos from more than 25 indoor environments. Our methods outperform
baseline approaches for traversability estimation on these new datasets.
|
TicTac: Accelerating Distributed Deep Learning with Communication
Scheduling | State-of-the-art deep learning systems rely on iterative distributed training
to tackle the increasing complexity of models and input data. The iteration
time in these communication-heavy systems depends on the computation time,
communication time and the extent of overlap of computation and communication.
In this work, we identify a shortcoming in systems with graph representation
for computation, such as TensorFlow and PyTorch, that result in high variance
in iteration time --- random order of received parameters across workers. We
develop a system, TicTac, to improve the iteration time by fixing this issue in
distributed deep learning with Parameter Servers while guaranteeing
near-optimal overlap of communication and computation. TicTac identifies and
enforces an order of network transfers which improves the iteration time using
prioritization. Our system is implemented over TensorFlow and requires no
changes to the model or developer inputs. TicTac improves the throughput by up
to $37.7\%$ in inference and $19.2\%$ in training, while also reducing
straggler effect by up to $2.3\times$. Our code is publicly available.
|
Deep Neural Network Compression with Single and Multiple Level
Quantization | Network quantization is an effective solution to compress deep neural
networks for practical usage. Existing network quantization methods cannot
sufficiently exploit the depth information to generate low-bit compressed
network. In this paper, we propose two novel network quantization approaches,
single-level network quantization (SLQ) for high-bit quantization and
multi-level network quantization (MLQ) for extremely low-bit quantization
(ternary).We are the first to consider the network quantization from both width
and depth level. In the width level, parameters are divided into two parts: one
for quantization and the other for re-training to eliminate the quantization
loss. SLQ leverages the distribution of the parameters to improve the width
level. In the depth level, we introduce incremental layer compensation to
quantize layers iteratively which decreases the quantization loss in each
iteration. The proposed approaches are validated with extensive experiments
based on the state-of-the-art neural networks including AlexNet, VGG-16,
GoogleNet and ResNet-18. Both SLQ and MLQ achieve impressive results.
|
Efficient Loss-Based Decoding on Graphs For Extreme Classification | In extreme classification problems, learning algorithms are required to map
instances to labels from an extremely large label set. We build on a recent
extreme classification framework with logarithmic time and space, and on a
general approach for error correcting output coding (ECOC) with loss-based
decoding, and introduce a flexible and efficient approach accompanied by
theoretical bounds. Our framework employs output codes induced by graphs, for
which we show how to perform efficient loss-based decoding to potentially
improve accuracy. In addition, our framework offers a tradeoff between
accuracy, model size and prediction time. We show how to find the sweet spot of
this tradeoff using only the training data. Our experimental study demonstrates
the validity of our assumptions and claims, and shows that our method is
competitive with state-of-the-art algorithms.
|
Learning Deep Generative Models of Graphs | Graphs are fundamental data structures which concisely capture the relational
structure in many important real-world domains, such as knowledge graphs,
physical and social interactions, language, and chemistry. Here we introduce a
powerful new approach for learning generative models over graphs, which can
capture both their structure and attributes. Our approach uses graph neural
networks to express probabilistic dependencies among a graph's nodes and edges,
and can, in principle, learn distributions over any arbitrary graph. In a
series of experiments our results show that once trained, our models can
generate good quality samples of both synthetic graphs as well as real
molecular graphs, both unconditionally and conditioned on data. Compared to
baselines that do not use graph-structured representations, our models often
perform far better. We also explore key challenges of learning generative
models of graphs, such as how to handle symmetries and ordering of elements
during the graph generation process, and offer possible solutions. Our work is
the first and most general approach for learning generative models over
arbitrary graphs, and opens new directions for moving away from restrictions of
vector- and sequence-like knowledge representations, toward more expressive and
flexible relational data structures.
|
Learning Approximate Inference Networks for Structured Prediction | Structured prediction energy networks (SPENs; Belanger & McCallum 2016) use
neural network architectures to define energy functions that can capture
arbitrary dependencies among parts of structured outputs. Prior work used
gradient descent for inference, relaxing the structured output to a set of
continuous variables and then optimizing the energy with respect to them. We
replace this use of gradient descent with a neural network trained to
approximate structured argmax inference. This "inference network" outputs
continuous values that we treat as the output structure. We develop
large-margin training criteria for joint training of the structured energy
function and inference network. On multi-label classification we report
speed-ups of 10-60x compared to (Belanger et al, 2017) while also improving
accuracy. For sequence labeling with simple structured energies, our approach
performs comparably to exact inference while being much faster at test time. We
then demonstrate improved accuracy by augmenting the energy with a "label
language model" that scores entire output label sequences, showing it can
improve handling of long-distance dependencies in part-of-speech tagging.
Finally, we show how inference networks can replace dynamic programming for
test-time inference in conditional random fields, suggestive for their general
use for fast inference in structured settings.
|
Fast Decoding in Sequence Models using Discrete Latent Variables | Autoregressive sequence models based on deep neural networks, such as RNNs,
Wavenet and the Transformer attain state-of-the-art results on many tasks.
However, they are difficult to parallelize and are thus slow at processing long
sequences. RNNs lack parallelism both during training and decoding, while
architectures like WaveNet and Transformer are much more parallelizable during
training, yet still operate sequentially during decoding.
Inspired by [arxiv:1711.00937], we present a method to extend sequence models
using discrete latent variables that makes decoding much more parallelizable.
We first auto-encode the target sequence into a shorter sequence of discrete
latent variables, which at inference time is generated autoregressively, and
finally decode the output sequence from this shorter latent sequence in
parallel. To this end, we introduce a novel method for constructing a sequence
of discrete latent variables and compare it with previously introduced methods.
Finally, we evaluate our model end-to-end on the task of neural machine
translation, where it is an order of magnitude faster at decoding than
comparable autoregressive models. While lower in BLEU than purely
autoregressive models, our model achieves higher scores than previously
proposed non-autoregressive translation models.
|
High-Accuracy Low-Precision Training | Low-precision computation is often used to lower the time and energy cost of
machine learning, and recently hardware accelerators have been developed to
support it. Still, it has been used primarily for inference - not training.
Previous low-precision training algorithms suffered from a fundamental
tradeoff: as the number of bits of precision is lowered, quantization noise is
added to the model, which limits statistical accuracy. To address this issue,
we describe a simple low-precision stochastic gradient descent variant called
HALP. HALP converges at the same theoretical rate as full-precision algorithms
despite the noise introduced by using low precision throughout execution. The
key idea is to use SVRG to reduce gradient variance, and to combine this with a
novel technique called bit centering to reduce quantization error. We show that
on the CPU, HALP can run up to $4 \times$ faster than full-precision SVRG and
can match its convergence trajectory. We implemented HALP in TensorQuant, and
show that it exceeds the validation performance of plain low-precision SGD on
two deep learning tasks.
|
RippleNet: Propagating User Preferences on the Knowledge Graph for
Recommender Systems | To address the sparsity and cold start problem of collaborative filtering,
researchers usually make use of side information, such as social networks or
item attributes, to improve recommendation performance. This paper considers
the knowledge graph as the source of side information. To address the
limitations of existing embedding-based and path-based methods for
knowledge-graph-aware recommendation, we propose Ripple Network, an end-to-end
framework that naturally incorporates the knowledge graph into recommender
systems. Similar to actual ripples propagating on the surface of water, Ripple
Network stimulates the propagation of user preferences over the set of
knowledge entities by automatically and iteratively extending a user's
potential interests along links in the knowledge graph. The multiple "ripples"
activated by a user's historically clicked items are thus superposed to form
the preference distribution of the user with respect to a candidate item, which
could be used for predicting the final clicking probability. Through extensive
experiments on real-world datasets, we demonstrate that Ripple Network achieves
substantial gains in a variety of scenarios, including movie, book and news
recommendation, over several state-of-the-art baselines.
|
Improved and Scalable Online Learning of Spatial Concepts and Language
Models with Mapping | We propose a novel online learning algorithm, called SpCoSLAM 2.0, for
spatial concepts and lexical acquisition with high accuracy and scalability.
Previously, we proposed SpCoSLAM as an online learning algorithm based on
unsupervised Bayesian probabilistic model that integrates multimodal place
categorization, lexical acquisition, and SLAM. However, our original algorithm
had limited estimation accuracy owing to the influence of the early stages of
learning, and increased computational complexity with added training data.
Therefore, we introduce techniques such as fixed-lag rejuvenation to reduce the
calculation time while maintaining an accuracy higher than that of the original
algorithm. The results show that, in terms of estimation accuracy, the proposed
algorithm exceeds the original algorithm and is comparable to batch learning.
In addition, the calculation time of the proposed algorithm does not depend on
the amount of training data and becomes constant for each step of the scalable
algorithm. Our approach will contribute to the realization of long-term spatial
language interactions between humans and robots.
|
Construction of neural networks for realization of localized deep
learning | The subject of deep learning has recently attracted users of machine learning
from various disciplines, including: medical diagnosis and bioinformatics,
financial market analysis and online advertisement, speech and handwriting
recognition, computer vision and natural language processing, time series
forecasting, and search engines. However, theoretical development of deep
learning is still at its infancy. The objective of this paper is to introduce a
deep neural network (also called deep-net) approach to localized manifold
learning, with each hidden layer endowed with a specific learning task. For the
purpose of illustrations, we only focus on deep-nets with three hidden layers,
with the first layer for dimensionality reduction, the second layer for bias
reduction, and the third layer for variance reduction. A feedback component
also designed to eliminate outliers. The main theoretical result in this paper
is the order $\mathcal O\left(m^{-2s/(2s+d)}\right)$ of approximation of the
regression function with regularity $s$, in terms of the number $m$ of sample
points, where the (unknown) manifold dimension $d$ replaces the dimension $D$
of the sampling (Euclidean) space for shallow nets.
|
Explaining Black-box Android Malware Detection | Machine-learning models have been recently used for detecting malicious
Android applications, reporting impressive performances on benchmark datasets,
even when trained only on features statically extracted from the application,
such as system calls and permissions. However, recent findings have highlighted
the fragility of such in-vitro evaluations with benchmark datasets, showing
that very few changes to the content of Android malware may suffice to evade
detection. How can we thus trust that a malware detector performing well on
benchmark data will continue to do so when deployed in an operating
environment? To mitigate this issue, the most popular Android malware detectors
use linear, explainable machine-learning models to easily identify the most
influential features contributing to each decision. In this work, we generalize
this approach to any black-box machine- learning model, by leveraging a
gradient-based approach to identify the most influential local features. This
enables using nonlinear models to potentially increase accuracy without
sacrificing interpretability of decisions. Our approach also highlights the
global characteristics learned by the model to discriminate between benign and
malware applications. Finally, as shown by our empirical analysis on a popular
Android malware detection task, it also helps identifying potential
vulnerabilities of linear and nonlinear models against adversarial
manipulations.
|
On Generation of Adversarial Examples using Convex Programming | It has been observed that deep learning architectures tend to make erroneous
decisions with high reliability for particularly designed adversarial
instances. In this work, we show that the perturbation analysis of these
architectures provides a framework for generating adversarial instances by
convex programming which, for classification tasks, is able to recover variants
of existing non-adaptive adversarial methods. The proposed framework can be
used for the design of adversarial noise under various desirable constraints
and different types of networks. Moreover, this framework is capable of
explaining various existing adversarial methods and can be used to derive new
algorithms as well. We make use of these results to obtain novel algorithms.
The experiments show the competitive performance of the obtained solutions, in
terms of fooling ratio, when benchmarked with well-known adversarial methods.
|
Hourly-Similarity Based Solar Forecasting Using Multi-Model Machine
Learning Blending | With the increasing penetration of solar power into power systems,
forecasting becomes critical in power system operations. In this paper, an
hourly-similarity (HS) based method is developed for 1-hour-ahead (1HA) global
horizontal irradiance (GHI) forecasting. This developed method utilizes diurnal
patterns, statistical distinctions between different hours, and hourly
similarities in solar data to improve the forecasting accuracy. The HS-based
method is built by training multiple two-layer multi-model forecasting
framework (MMFF) models independently with the same-hour subsets. The final
optimal model is a combination of MMFF models with the best-performed blending
algorithm at every hour. At the forecasting stage, the most suitable model is
selected to perform the forecasting subtask of a certain hour. The HS-based
method is validated by 1-year data with six solar features collected by the
National Renewable Energy Laboratory (NREL). Results show that the HS-based
method outperforms the non-HS (all-in-one) method significantly with the same
MMFF architecture, wherein the optimal HS- based method outperforms the best
all-in-one method by 10.94% and 7.74% based on the normalized mean absolute
error and normalized root mean square error, respectively.
|
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks | Neural network pruning techniques can reduce the parameter counts of trained
networks by over 90%, decreasing storage requirements and improving
computational performance of inference without compromising accuracy. However,
contemporary experience is that the sparse architectures produced by pruning
are difficult to train from the start, which would similarly improve training
performance.
We find that a standard pruning technique naturally uncovers subnetworks
whose initializations made them capable of training effectively. Based on these
results, we articulate the "lottery ticket hypothesis:" dense,
randomly-initialized, feed-forward networks contain subnetworks ("winning
tickets") that - when trained in isolation - reach test accuracy comparable to
the original network in a similar number of iterations. The winning tickets we
find have won the initialization lottery: their connections have initial
weights that make training particularly effective.
We present an algorithm to identify winning tickets and a series of
experiments that support the lottery ticket hypothesis and the importance of
these fortuitous initializations. We consistently find winning tickets that are
less than 10-20% of the size of several fully-connected and convolutional
feed-forward architectures for MNIST and CIFAR10. Above this size, the winning
tickets that we find learn faster than the original network and reach higher
test accuracy.
|
Precision and Recall for Time Series | Classical anomaly detection is principally concerned with point-based
anomalies, those anomalies that occur at a single point in time. Yet, many
real-world anomalies are range-based, meaning they occur over a period of time.
Motivated by this observation, we present a new mathematical model to evaluate
the accuracy of time series classification algorithms. Our model expands the
well-known Precision and Recall metrics to measure ranges, while simultaneously
enabling customization support for domain-specific preferences.
|
Deep Auxiliary Learning for Visual Localization and Odometry | Localization is an indispensable component of a robot's autonomy stack that
enables it to determine where it is in the environment, essentially making it a
precursor for any action execution or planning. Although convolutional neural
networks have shown promising results for visual localization, they are still
grossly outperformed by state-of-the-art local feature-based techniques. In
this work, we propose VLocNet, a new convolutional neural network architecture
for 6-DoF global pose regression and odometry estimation from consecutive
monocular images. Our multitask model incorporates hard parameter sharing, thus
being compact and enabling real-time inference, in addition to being end-to-end
trainable. We propose a novel loss function that utilizes auxiliary learning to
leverage relative pose information during training, thereby constraining the
search space to obtain consistent pose estimates. We evaluate our proposed
VLocNet on indoor as well as outdoor datasets and show that even our single
task model exceeds the performance of state-of-the-art deep architectures for
global localization, while achieving competitive performance for visual
odometry estimation. Furthermore, we present extensive experimental evaluations
utilizing our proposed Geometric Consistency Loss that show the effectiveness
of multitask learning and demonstrate that our model is the first deep learning
technique to be on par with, and in some cases outperforms state-of-the-art
SIFT-based approaches.
|
Syntax-Aware Language Modeling with Recurrent Neural Networks | Neural language models (LMs) are typically trained using only lexical
features, such as surface forms of words. In this paper, we argue this deprives
the LM of crucial syntactic signals that can be detected at high confidence
using existing parsers. We present a simple but highly effective approach for
training neural LMs using both lexical and syntactic information, and a novel
approach for applying such LMs to unparsed text using sequential Monte Carlo
sampling. In experiments on a range of corpora and corpus sizes, we show our
approach consistently outperforms standard lexical LMs in character-level
language modeling; on the other hand, for word-level models the models are on a
par with standard language models. These results indicate potential for
expanding LMs beyond lexical surface features to higher-level NLP features for
character-level models.
|
Standing Wave Decomposition Gaussian Process | We propose a Standing Wave Decomposition (SWD) approximation to Gaussian
Process regression (GP). GP involves a costly matrix inversion operation, which
limits applicability to large data analysis. For an input space that can be
approximated by a grid and when correlations among data are short-ranged, the
kernel matrix inversion can be replaced by analytic diagonalization using the
SWD. We show that this approach applies to uni- and multi-dimensional input
data, extends to include longer-range correlations, and the grid can be in a
latent space and used as inducing points. Through simulations, we show that our
approximate method applied to the squared exponential kernel outperforms
existing methods in predictive accuracy per unit time in the regime where data
are plentiful. Our SWD-GP is recommended for regression analyses where there is
a relatively large amount of data and/or there are constraints on computation
time.
|
Competitive Machine Learning: Best Theoretical Prediction vs
Optimization | Machine learning is often used in competitive scenarios: Participants learn
and fit static models, and those models compete in a shared platform. The
common assumption is that in order to win a competition one has to have the
best predictive model, i.e., the model with the smallest out-sample error. Is
that necessarily true? Does the best theoretical predictive model for a target
always yield the best reward in a competition? If not, can one take the best
model and purposefully change it into a theoretically inferior model which in
practice results in a higher competitive edge? How does that modification look
like? And finally, if all participants modify their prediction models towards
the best practical performance, who benefits the most? players with inferior
models, or those with theoretical superiority? The main theme of this paper is
to raise these important questions and propose a theoretical model to answer
them. We consider a study case where two linear predictive models compete over
a shared target. The model with the closest estimate gets the whole reward,
which is equal to the absolute value of the target. We characterize the reward
function of each model, and using a basic game theoretic approach, demonstrate
that the inferior competitor can significantly improve his performance by
choosing optimal model coefficients that are different from the best
theoretical prediction. This is a preliminary study that emphasizes the fact
that in many applications where predictive machine learning is at the service
of competition, much can be gained from practical (back-testing) optimization
of the model compared to static prediction improvement.
|
Sequential Outlier Detection based on Incremental Decision Trees | We introduce an online outlier detection algorithm to detect outliers in a
sequentially observed data stream. For this purpose, we use a two-stage
filtering and hedging approach. In the first stage, we construct a multi-modal
probability density function to model the normal samples. In the second stage,
given a new observation, we label it as an anomaly if the value of
aforementioned density function is below a specified threshold at the newly
observed point. In order to construct our multi-modal density function, we use
an incremental decision tree to construct a set of subspaces of the observation
space. We train a single component density function of the exponential family
using the observations, which fall inside each subspace represented on the
tree. These single component density functions are then adaptively combined to
produce our multi-modal density function, which is shown to achieve the
performance of the best convex combination of the density functions defined on
the subspaces. As we observe more samples, our tree grows and produces more
subspaces. As a result, our modeling power increases in time, while mitigating
overfitting issues. In order to choose our threshold level to label the
observations, we use an adaptive thresholding scheme. We show that our adaptive
threshold level achieves the performance of the optimal pre-fixed threshold
level, which knows the observation labels in hindsight. Our algorithm provides
significant performance improvements over the state of the art in our wide set
of experiments involving both synthetic as well as real data.
|
Scoring Formulation for Multi-Condition Joint PLDA | The joint PLDA model, is a generalization of PLDA where the nuisance variable
is no longer considered independent across samples, but potentially shared
(tied) across samples that correspond to the same nuisance condition. The
original work considered a single nuisance condition, deriving the EM and
scoring formulas for this scenario. In this document, we show how to obtain
likelihood ratios for scoring when multiple nuisance conditions are allowed in
the model.
|
DeepMoTIon: Learning to Navigate Like Humans | We present a novel human-aware navigation approach, where the robot learns to
mimic humans to navigate safely in crowds. The presented model, referred to as
DeepMoTIon, is trained with pedestrian surveillance data to predict human
velocity in the environment. The robot processes LiDAR scans via the trained
network to navigate to the target location. We conduct extensive experiments to
assess the components of our network and prove their necessity to imitate
humans. Our experiments show that DeepMoTIion outperforms all the benchmarks in
terms of human imitation, achieving a 24% reduction in time series-based path
deviation over the next best approach. In addition, while many other approaches
often failed to reach the target, our method reached the target in 100% of the
test cases while complying with social norms and ensuring human safety.
|
Attention-based Graph Neural Network for Semi-supervised Learning | Recently popularized graph neural networks achieve the state-of-the-art
accuracy on a number of standard benchmark datasets for graph-based
semi-supervised learning, improving significantly over existing approaches.
These architectures alternate between a propagation layer that aggregates the
hidden states of the local neighborhood and a fully-connected layer. Perhaps
surprisingly, we show that a linear model, that removes all the intermediate
fully-connected layers, is still able to achieve a performance comparable to
the state-of-the-art models. This significantly reduces the number of
parameters, which is critical for semi-supervised learning where number of
labeled examples are small. This in turn allows a room for designing more
innovative propagation layers. Based on this insight, we propose a novel graph
neural network that removes all the intermediate fully-connected layers, and
replaces the propagation layers with attention mechanisms that respect the
structure of the graph. The attention mechanism allows us to learn a dynamic
and adaptive local summary of the neighborhood to achieve more accurate
predictions. In a number of experiments on benchmark citation networks
datasets, we demonstrate that our approach outperforms competing methods. By
examining the attention weights among neighbors, we show that our model
provides some interesting insights on how neighbors influence each other.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.