title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Learning Dynamic Classes of Events using Stacked Multilayer Perceptron
Networks | People often use a web search engine to find information about events of
interest, for example, sport competitions, political elections, festivals and
entertainment news. In this paper, we study a problem of detecting
event-related queries, which is the first step before selecting a suitable
time-aware retrieval model. In general, event-related information needs can be
observed in query streams through various temporal patterns of user search
behavior, e.g., spiky peaks for popular events, and periodicities for
repetitive events. However, it is also common that users search for non-popular
events, which may not exhibit temporal variations in query streams, e.g., past
events recently occurred, historical events triggered by anniversaries or
similar events, and future events anticipated to happen. To address the
challenge of detecting dynamic classes of events, we propose a novel deep
learning model to classify a given query into a predetermined set of multiple
event types. Our proposed model, a Stacked Multilayer Perceptron (S-MLP)
network, consists of multilayer perceptron used as a basic learning unit. We
assemble stacked units to further learn complex relationships between neutrons
in successive layers. To evaluate our proposed model, we conduct experiments
using real-world queries and a set of manually created ground truth.
Preliminary results have shown that our proposed deep learning model
outperforms the state-of-the-art classification models significantly.
|
Deep Learning Markov Random Field for Semantic Segmentation | Semantic segmentation tasks can be well modeled by Markov Random Field (MRF).
This paper addresses semantic segmentation by incorporating high-order
relations and mixture of label contexts into MRF. Unlike previous works that
optimized MRFs using iterative algorithm, we solve MRF by proposing a
Convolutional Neural Network (CNN), namely Deep Parsing Network (DPN), which
enables deterministic end-to-end computation in a single forward pass.
Specifically, DPN extends a contemporary CNN to model unary terms and
additional layers are devised to approximate the mean field (MF) algorithm for
pairwise terms. It has several appealing properties. First, different from the
recent works that required many iterations of MF during back-propagation, DPN
is able to achieve high performance by approximating one iteration of MF.
Second, DPN represents various types of pairwise terms, making many existing
models as its special cases. Furthermore, pairwise terms in DPN provide a
unified framework to encode rich contextual information in high-dimensional
data, such as images and videos. Third, DPN makes MF easier to be parallelized
and speeded up, thus enabling efficient inference. DPN is thoroughly evaluated
on standard semantic image/video segmentation benchmarks, where a single DPN
model yields state-of-the-art segmentation accuracies on PASCAL VOC 2012,
Cityscapes dataset and CamVid dataset.
|
Algorithmic Composition of Melodies with Deep Recurrent Neural Networks | A big challenge in algorithmic composition is to devise a model that is both
easily trainable and able to reproduce the long-range temporal dependencies
typical of music. Here we investigate how artificial neural networks can be
trained on a large corpus of melodies and turned into automated music composers
able to generate new melodies coherent with the style they have been trained
on. We employ gated recurrent unit networks that have been shown to be
particularly efficient in learning complex sequential activations with
arbitrary long time lags. Our model processes rhythm and melody in parallel
while modeling the relation between these two features. Using such an approach,
we were able to generate interesting complete melodies or suggest possible
continuations of a melody fragment that is coherent with the characteristics of
the fragment itself.
|
On the Theoretical Capacity of Evolution Strategies to Statistically
Learn the Landscape Hessian | We study the theoretical capacity to statistically learn local landscape
information by Evolution Strategies (ESs). Specifically, we investigate the
covariance matrix when constructed by ESs operating with the selection operator
alone. We model continuous generation of candidate solutions about quadratic
basins of attraction, with deterministic selection of the decision vectors that
minimize the objective function values. Our goal is to rigorously show that
accumulation of winning individuals carries the potential to reveal valuable
information about the search landscape, e.g., as already practically utilized
by derandomized ES variants. We first show that the statistically-constructed
covariance matrix over such winning decision vectors shares the same
eigenvectors with the Hessian matrix about the optimum. We then provide an
analytic approximation of this covariance matrix for a non-elitist multi-child
$(1,\lambda)$-strategy, which holds for a large population size $\lambda$.
Finally, we also numerically corroborate our results.
|
Multiclass feature learning for hyperspectral image classification:
sparse and hierarchical solutions | In this paper, we tackle the question of discovering an effective set of
spatial filters to solve hyperspectral classification problems. Instead of
fixing a priori the filters and their parameters using expert knowledge, we let
the model find them within random draws in the (possibly infinite) space of
possible filters. We define an active set feature learner that includes in the
model only features that improve the classifier. To this end, we consider a
fast and linear classifier, multiclass logistic classification, and show that
with a good representation (the filters discovered), such a simple classifier
can reach at least state of the art performances. We apply the proposed active
set learner in four hyperspectral image classification problems, including
agricultural and urban classification at different resolutions, as well as
multimodal data. We also propose a hierarchical setting, which allows to
generate more complex banks of features that can better describe the
nonlinearities present in the data.
|
Event Abstraction for Process Mining using Supervised Learning
Techniques | Process mining techniques focus on extracting insight in processes from event
logs. In many cases, events recorded in the event log are too fine-grained,
causing process discovery algorithms to discover incomprehensible process
models or process models that are not representative of the event log. We show
that when process discovery algorithms are only able to discover an
unrepresentative process model from a low-level event log, structure in the
process can in some cases still be discovered by first abstracting the event
log to a higher level of granularity. This gives rise to the challenge to
bridge the gap between an original low-level event log and a desired high-level
perspective on this log, such that a more structured or more comprehensible
process model can be discovered. We show that supervised learning can be
leveraged for the event abstraction task when annotations with high-level
interpretations of the low-level events are available for a subset of the
sequences (i.e., traces). We present a method to generate feature vector
representations of events based on XES extensions, and describe an approach to
abstract events in an event log with Condition Random Fields using these event
features. Furthermore, we propose a sequence-focused metric to evaluate
supervised event abstraction results that fits closely to the tasks of process
discovery and conformance checking. We conclude this paper by demonstrating the
usefulness of supervised event abstraction for obtaining more structured and/or
more comprehensible process models using both real life event data and
synthetic event data.
|
Importance sampling strategy for non-convex randomized block-coordinate
descent | As the number of samples and dimensionality of optimization problems related
to statistics an machine learning explode, block coordinate descent algorithms
have gained popularity since they reduce the original problem to several
smaller ones. Coordinates to be optimized are usually selected randomly
according to a given probability distribution. We introduce an importance
sampling strategy that helps randomized coordinate descent algorithms to focus
on blocks that are still far from convergence. The framework applies to
problems composed of the sum of two possibly non-convex terms, one being
separable and non-smooth. We have compared our algorithm to a full gradient
proximal approach as well as to a randomized block coordinate algorithm that
considers uniform sampling and cyclic block coordinate descent. Experimental
evidences show the clear benefit of using an importance sampling strategy.
|
Non-convex regularization in remote sensing | In this paper, we study the effect of different regularizers and their
implications in high dimensional image classification and sparse linear
unmixing. Although kernelization or sparse methods are globally accepted
solutions for processing data in high dimensions, we present here a study on
the impact of the form of regularization used and its parametrization. We
consider regularization via traditional squared (2) and sparsity-promoting (1)
norms, as well as more unconventional nonconvex regularizers (p and Log Sum
Penalty). We compare their properties and advantages on several classification
and linear unmixing tasks and provide advices on the choice of the best
regularizer for the problem at hand. Finally, we also provide a fully
functional toolbox for the community.
|
Explaining Predictions of Non-Linear Classifiers in NLP | Layer-wise relevance propagation (LRP) is a recently proposed technique for
explaining predictions of complex non-linear classifiers in terms of input
variables. In this paper, we apply LRP for the first time to natural language
processing (NLP). More precisely, we use it to explain the predictions of a
convolutional neural network (CNN) trained on a topic categorization task. Our
analysis highlights which words are relevant for a specific prediction of the
CNN. We compare our technique to standard sensitivity analysis, both
qualitatively and quantitatively, using a "word deleting" perturbation
experiment, a PCA analysis, and various visualizations. All experiments
validate the suitability of LRP for explaining the CNN predictions, which is
also in line with results reported in recent image classification studies.
|
Unsupervised preprocessing for Tactile Data | Tactile information is important for gripping, stable grasp, and in-hand
manipulation, yet the complexity of tactile data prevents widespread use of
such sensors. We make use of an unsupervised learning algorithm that transforms
the complex tactile data into a compact, latent representation without the need
to record ground truth reference data. These compact representations can either
be used directly in a reinforcement learning based controller or can be used to
calibrate the tactile sensor to physical quantities with only a few datapoints.
We show the quality of our latent representation by predicting important
features and with a simple control task.
|
Nearly-optimal Robust Matrix Completion | In this paper, we consider the problem of Robust Matrix Completion (RMC)
where the goal is to recover a low-rank matrix by observing a small number of
its entries out of which a few can be arbitrarily corrupted. We propose a
simple projected gradient descent method to estimate the low-rank matrix that
alternately performs a projected gradient descent step and cleans up a few of
the corrupted entries using hard-thresholding. Our algorithm solves RMC using
nearly optimal number of observations as well as nearly optimal number of
corruptions. Our result also implies significant improvement over the existing
time complexity bounds for the low-rank matrix completion problem. Finally, an
application of our result to the robust PCA problem (low-rank+sparse matrix
separation) leads to nearly linear time (in matrix dimensions) algorithm for
the same; existing state-of-the-art methods require quadratic time. Our
empirical results corroborate our theoretical results and show that even for
moderate sized problems, our method for robust PCA is an an order of magnitude
faster than the existing methods.
|
DropNeuron: Simplifying the Structure of Deep Neural Networks | Deep learning using multi-layer neural networks (NNs) architecture manifests
superb power in modern machine learning systems. The trained Deep Neural
Networks (DNNs) are typically large. The question we would like to address is
whether it is possible to simplify the NN during training process to achieve a
reasonable performance within an acceptable computational time. We presented a
novel approach of optimising a deep neural network through regularisation of
net- work architecture. We proposed regularisers which support a simple
mechanism of dropping neurons during a network training process. The method
supports the construction of a simpler deep neural networks with compatible
performance with its simplified version. As a proof of concept, we evaluate the
proposed method with examples including sparse linear regression, deep
autoencoder and convolutional neural network. The valuations demonstrate
excellent performance.
The code for this work can be found in
http://www.github.com/panweihit/DropNeuron
|
Analyzing the Behavior of Visual Question Answering Models | Recently, a number of deep-learning based models have been proposed for the
task of Visual Question Answering (VQA). The performance of most models is
clustered around 60-70%. In this paper we propose systematic methods to analyze
the behavior of these models as a first step towards recognizing their
strengths and weaknesses, and identifying the most fruitful directions for
progress. We analyze two models, one each from two major classes of VQA models
-- with-attention and without-attention and show the similarities and
differences in the behavior of these models. We also analyze the winning entry
of the VQA Challenge 2016.
Our behavior analysis reveals that despite recent progress, today's VQA
models are "myopic" (tend to fail on sufficiently novel instances), often "jump
to conclusions" (converge on a predicted answer after 'listening' to just half
the question), and are "stubborn" (do not change their answers across images).
|
Parallel SGD: When does averaging help? | Consider a number of workers running SGD independently on the same pool of
data and averaging the models every once in a while -- a common but not well
understood practice. We study model averaging as a variance-reducing mechanism
and describe two ways in which the frequency of averaging affects convergence.
For convex objectives, we show the benefit of frequent averaging depends on the
gradient variance envelope. For non-convex objectives, we illustrate that this
benefit depends on the presence of multiple globally optimal points. We
complement our findings with multicore experiments on both synthetic and real
data.
|
Personalized Prognostic Models for Oncology: A Machine Learning Approach | We have applied a little-known data transformation to subsets of the
Surveillance, Epidemiology, and End Results (SEER) publically available data of
the National Cancer Institute (NCI) to make it suitable input to standard
machine learning classifiers. This transformation properly treats the
right-censored data in the SEER data and the resulting Random Forest and
Multi-Layer Perceptron models predict full survival curves. Treating the 6, 12,
and 60 months points of the resulting survival curves as 3 binary classifiers,
the 18 resulting classifiers have AUC values ranging from .765 to .885. Further
evidence that the models have generalized well from the training data is
provided by the extremely high levels of agreement between the random forest
and neural network models predictions on the 6, 12, and 60 month binary
classifiers.
|
Multi-Stage Temporal Difference Learning for 2048-like Games | Szubert and Jaskowski successfully used temporal difference (TD) learning
together with n-tuple networks for playing the game 2048. However, we observed
a phenomenon that the programs based on TD learning still hardly reach large
tiles. In this paper, we propose multi-stage TD (MS-TD) learning, a kind of
hierarchical reinforcement learning method, to effectively improve the
performance for the rates of reaching large tiles, which are good metrics to
analyze the strength of 2048 programs. Our experiments showed significant
improvements over the one without using MS-TD learning. Namely, using 3-ply
expectimax search, the program with MS-TD learning reached 32768-tiles with a
rate of 18.31%, while the one with TD learning did not reach any. After further
tuned, our 2048 program reached 32768-tiles with a rate of 31.75% in 10,000
games, and one among these games even reached a 65536-tile, which is the first
ever reaching a 65536-tile to our knowledge. In addition, MS-TD learning method
can be easily applied to other 2048-like games, such as Threes. Based on MS-TD
learning, our experiments for Threes also demonstrated similar performance
improvement, where the program with MS-TD learning reached 6144-tiles with a
rate of 7.83%, while the one with TD learning only reached 0.45%.
|
Robust Learning of Fixed-Structure Bayesian Networks | We investigate the problem of learning Bayesian networks in a robust model
where an $\epsilon$-fraction of the samples are adversarially corrupted. In
this work, we study the fully observable discrete case where the structure of
the network is given. Even in this basic setting, previous learning algorithms
either run in exponential time or lose dimension-dependent factors in their
error guarantees. We provide the first computationally efficient robust
learning algorithm for this problem with dimension-independent error
guarantees. Our algorithm has near-optimal sample complexity, runs in
polynomial time, and achieves error that scales nearly-linearly with the
fraction of adversarially corrupted samples. Finally, we show on both synthetic
and semi-synthetic data that our algorithm performs well in practice.
|
Deep Recurrent Neural Networks for Supernovae Classification | We apply deep recurrent neural networks, which are capable of learning
complex sequential information, to classify supernovae\footnote{Code available
at
\href{https://github.com/adammoss/supernovae}{https://github.com/adammoss/supernovae}}.
The observational time and filter fluxes are used as inputs to the network, but
since the inputs are agnostic additional data such as host galaxy information
can also be included. Using the Supernovae Photometric Classification Challenge
(SPCC) data, we find that deep networks are capable of learning about light
curves, however the performance of the network is highly sensitive to the
amount of training data. For a training size of 50\% of the representational
SPCC dataset (around $10^4$ supernovae) we obtain a type-Ia vs. non-type-Ia
classification accuracy of 94.7\%, an area under the Receiver Operating
Characteristic curve AUC of 0.986 and a SPCC figure-of-merit $F_1=0.64$. When
using only the data for the early-epoch challenge defined by the SPCC we
achieve a classification accuracy of 93.1\%, AUC of 0.977 and $F_1=0.58$,
results almost as good as with the whole light-curve. By employing
bidirectional neural networks we can acquire impressive classification results
between supernovae types -I,~-II and~-III at an accuracy of 90.4\% and AUC of
0.974. We also apply a pre-trained model to obtain classification probabilities
as a function of time, and show it can give early indications of supernovae
type. Our method is competitive with existing algorithms and has applications
for future large-scale photometric surveys.
|
The VGLC: The Video Game Level Corpus | Levels are a key component of many different video games, and a large body of
work has been produced on how to procedurally generate game levels. Recently,
Machine Learning techniques have been applied to video game level generation
towards the purpose of automatically generating levels that have the properties
of the training corpus. Towards that end we have made available a corpora of
video game levels in an easy to parse format ideal for different machine
learning and other game AI research purposes.
|
Sort Story: Sorting Jumbled Images and Captions into Stories | Temporal common sense has applications in AI tasks such as QA, multi-document
summarization, and human-AI communication. We propose the task of sequencing --
given a jumbled set of aligned image-caption pairs that belong to a story, the
task is to sort them such that the output sequence forms a coherent story. We
present multiple approaches, via unary (position) and pairwise (order)
predictions, and their ensemble-based combinations, achieving strong results on
this task. We use both text-based and image-based features, which depict
complementary improvements. Using qualitative examples, we demonstrate that our
models have learnt interesting aspects of temporal common sense.
|
Is a Picture Worth Ten Thousand Words in a Review Dataset? | While textual reviews have become prominent in many recommendation-based
systems, automated frameworks to provide relevant visual cues against text
reviews where pictures are not available is a new form of task confronted by
data mining and machine learning researchers. Suggestions of pictures that are
relevant to the content of a review could significantly benefit the users by
increasing the effectiveness of a review. We propose a deep learning-based
framework to automatically: (1) tag the images available in a review dataset,
(2) generate a caption for each image that does not have one, and (3) enhance
each review by recommending relevant images that might not be uploaded by the
corresponding reviewer. We evaluate the proposed framework using the Yelp
Challenge Dataset. While a subset of the images in this particular dataset are
correctly captioned, the majority of the pictures do not have any associated
text. Moreover, there is no mapping between reviews and images. Each image has
a corresponding business-tag where the picture was taken, though. The overall
data setting and unavailability of crucial pieces required for a mapping make
the problem of recommending images for reviews a major challenge. Qualitative
and quantitative evaluations indicate that our proposed framework provides high
quality enhancements through automatic captioning, tagging, and recommendation
for mapping reviews and images.
|
On the Solvability of Inductive Problems: A Study in Epistemic Topology | We investigate the issues of inductive problem-solving and learning by
doxastic agents. We provide topological characterizations of solvability and
learnability, and we use them to prove that AGM-style belief revision is
"universal", i.e., that every solvable problem is solvable by AGM conditioning.
|
Satisfying Real-world Goals with Dataset Constraints | The goal of minimizing misclassification error on a training set is often
just one of several real-world goals that might be defined on different
datasets. For example, one may require a classifier to also make positive
predictions at some specified rate for some subpopulation (fairness), or to
achieve a specified empirical recall. Other real-world goals include reducing
churn with respect to a previously deployed model, or stabilizing online
training. In this paper we propose handling multiple goals on multiple datasets
by training with dataset constraints, using the ramp penalty to accurately
quantify costs, and present an efficient algorithm to approximately optimize
the resulting non-convex constrained optimization problem. Experiments on both
benchmark and real-world industry datasets demonstrate the effectiveness of our
approach.
|
Multipartite Ranking-Selection of Low-Dimensional Instances by
Supervised Projection to High-Dimensional Space | Pruning of redundant or irrelevant instances of data is a key to every
successful solution for pattern recognition. In this paper, we present a novel
ranking-selection framework for low-length but highly correlated instances.
Instead of working in the low-dimensional instance space, we learn a supervised
projection to high-dimensional space spanned by the number of classes in the
dataset under study. Imposing higher distinctions via exposing the notion of
labels to the instances, lets to deploy one versus all ranking for each
individual classes and selecting quality instances via adaptive thresholding of
the overall scores. To prove the efficiency of our paradigm, we employ it for
the purpose of texture understanding which is a hard recognition challenge due
to high similarity of texture pixels and low dimensionality of their color
features. Our experiments show considerable improvements in recognition
performance over other local descriptors on several publicly available
datasets.
|
Regression Trees and Random forest based feature selection for malaria
risk exposure prediction | This paper deals with prediction of anopheles number, the main vector of
malaria risk, using environmental and climate variables. The variables
selection is based on an automatic machine learning method using regression
trees, and random forests combined with stratified two levels cross validation.
The minimum threshold of variables importance is accessed using the quadratic
distance of variables importance while the optimal subset of selected variables
is used to perform predictions. Finally the results revealed to be
qualitatively better, at the selection, the prediction , and the CPU time point
of view than those obtained by GLM-Lasso method.
|
Is the Bellman residual a bad proxy? | This paper aims at theoretically and empirically comparing two standard
optimization criteria for Reinforcement Learning: i) maximization of the mean
value and ii) minimization of the Bellman residual. For that purpose, we place
ourselves in the framework of policy search algorithms, that are usually
designed to maximize the mean value, and derive a method that minimizes the
residual $\|T_* v_\pi - v_\pi\|_{1,\nu}$ over policies. A theoretical analysis
shows how good this proxy is to policy optimization, and notably that it is
better than its value-based counterpart. We also propose experiments on
randomly generated generic Markov decision processes, specifically designed for
studying the influence of the involved concentrability coefficient. They show
that the Bellman residual is generally a bad proxy to policy optimization and
that directly maximizing the mean value is much better, despite the current
lack of deep theoretical analysis. This might seem obvious, as directly
addressing the problem of interest is usually better, but given the prevalence
of (projected) Bellman residual minimization in value-based reinforcement
learning, we believe that this question is worth to be considered.
|
Hybrid Recommender System based on Autoencoders | A standard model for Recommender Systems is the Matrix Completion setting:
given partially known matrix of ratings given by users (rows) to items
(columns), infer the unknown ratings. In the last decades, few attempts where
done to handle that objective with Neural Networks, but recently an
architecture based on Autoencoders proved to be a promising approach. In
current paper, we enhanced that architecture (i) by using a loss function
adapted to input data with missing values, and (ii) by incorporating side
information. The experiments demonstrate that while side information only
slightly improve the test error averaged on all users/items, it has more impact
on cold users/items.
|
Collective Semi-Supervised Learning for User Profiling in Social Media | The abundance of user-generated data in social media has incentivized the
development of methods to infer the latent attributes of users, which are
crucially useful for personalization, advertising and recommendation. However,
the current user profiling approaches have limited success, due to the lack of
a principled way to integrate different types of social relationships of a
user, and the reliance on scarcely-available labeled data in building a
prediction model. In this paper, we present a novel solution termed Collective
Semi-Supervised Learning (CSL), which provides a principled means to integrate
different types of social relationship and unlabeled data under a unified
computational framework. The joint learning from multiple relationships and
unlabeled data yields a computationally sound and accurate approach to model
user attributes in social media. Extensive experiments using Twitter data have
demonstrated the efficacy of our CSL approach in inferring user attributes such
as account type and marital status. We also show how CSL can be used to
determine important user features, and to make inference on a larger user
population.
|
Neural Network Based Next-Song Recommendation | Recently, the next-item/basket recommendation system, which considers the
sequential relation between bought items, has drawn attention of researchers.
The utilization of sequential patterns has boosted performance on several kinds
of recommendation tasks. Inspired by natural language processing (NLP)
techniques, we propose a novel neural network (NN) based next-song recommender,
CNN-rec, in this paper. Then, we compare the proposed system with several NN
based and classic recommendation systems on the next-song recommendation task.
Verification results indicate the proposed system outperforms classic systems
and has comparable performance with the state-of-the-art system.
|
Sampling-based Gradient Regularization for Capturing Long-Term
Dependencies in Recurrent Neural Networks | Vanishing (and exploding) gradients effect is a common problem for recurrent
neural networks with nonlinear activation functions which use backpropagation
method for calculation of derivatives. Deep feedforward neural networks with
many hidden layers also suffer from this effect. In this paper we propose a
novel universal technique that makes the norm of the gradient stay in the
suitable range. We construct a way to estimate a contribution of each training
example to the norm of the long-term components of the target function s
gradient. Using this subroutine we can construct mini-batches for the
stochastic gradient descent (SGD) training that leads to high performance and
accuracy of the trained network even for very complex tasks. We provide a
straightforward mathematical estimation of minibatch s impact on for the
gradient norm and prove its correctness theoretically. To check our framework
experimentally we use some special synthetic benchmarks for testing RNNs on
ability to capture long-term dependencies. Our network can detect links between
events in the (temporal) sequence at the range approx. 100 and longer.
|
Precise neural network computation with imprecise analog devices | The operations used for neural network computation map favorably onto simple
analog circuits, which outshine their digital counterparts in terms of
compactness and efficiency. Nevertheless, such implementations have been
largely supplanted by digital designs, partly because of device mismatch
effects due to material and fabrication imperfections. We propose a framework
that exploits the power of deep learning to compensate for this mismatch by
incorporating the measured device variations as constraints in the neural
network training process. This eliminates the need for mismatch minimization
strategies and allows circuit complexity and power-consumption to be reduced to
a minimum. Our results, based on large-scale simulations as well as a prototype
VLSI chip implementation indicate a processing efficiency comparable to current
state-of-art digital implementations. This method is suitable for future
technology based on nanodevices with large variability, such as memristive
arrays.
|
Wide & Deep Learning for Recommender Systems | Generalized linear models with nonlinear feature transformations are widely
used for large-scale regression and classification problems with sparse inputs.
Memorization of feature interactions through a wide set of cross-product
feature transformations are effective and interpretable, while generalization
requires more feature engineering effort. With less feature engineering, deep
neural networks can generalize better to unseen feature combinations through
low-dimensional dense embeddings learned for the sparse features. However, deep
neural networks with embeddings can over-generalize and recommend less relevant
items when the user-item interactions are sparse and high-rank. In this paper,
we present Wide & Deep learning---jointly trained wide linear models and deep
neural networks---to combine the benefits of memorization and generalization
for recommender systems. We productionized and evaluated the system on Google
Play, a commercial mobile app store with over one billion active users and over
one million apps. Online experiment results show that Wide & Deep significantly
increased app acquisitions compared with wide-only and deep-only models. We
have also open-sourced our implementation in TensorFlow.
|
Sequence-Level Knowledge Distillation | Neural machine translation (NMT) offers a novel alternative formulation of
translation that is potentially simpler than statistical approaches. However to
reach competitive performance, NMT models need to be exceedingly large. In this
paper we consider applying knowledge distillation approaches (Bucila et al.,
2006; Hinton et al., 2015) that have proven successful for reducing the size of
neural models in other domains to the problem of NMT. We demonstrate that
standard knowledge distillation applied to word-level prediction can be
effective for NMT, and also introduce two novel sequence-level versions of
knowledge distillation that further improve performance, and somewhat
surprisingly, seem to eliminate the need for beam search (even when applied on
the original teacher model). Our best student model runs 10 times faster than
its state-of-the-art teacher with little loss in performance. It is also
significantly better than a baseline model trained without knowledge
distillation: by 4.2/1.7 BLEU with greedy decoding/beam search. Applying weight
pruning on top of knowledge distillation results in a student model that has 13
times fewer parameters than the original teacher model, with a decrease of 0.4
BLEU.
|
Bidirectional Recurrent Neural Networks for Medical Event Detection in
Electronic Health Records | Sequence labeling for extraction of medical events and their attributes from
unstructured text in Electronic Health Record (EHR) notes is a key step towards
semantic understanding of EHRs. It has important applications in health
informatics including pharmacovigilance and drug surveillance. The state of the
art supervised machine learning models in this domain are based on Conditional
Random Fields (CRFs) with features calculated from fixed context windows. In
this application, we explored various recurrent neural network frameworks and
show that they significantly outperformed the CRF models.
|
Gear fault diagnosis based on Gaussian correlation of vibrations signals
and wavelet coefficients | The features of non-stationary multi-component signals are often difficult to
be extracted for expert systems. In this paper, a new method for feature
extraction that is based on maximization of local Gaussian correlation function
of wavelet coefficients and signal is presented. The effect of empirical mode
decomposition (EMD) to decompose multi-component signals to intrinsic mode
functions (IMFs), before using of local Gaussian correlation is discussed. The
experimental vibration signals from two gearbox systems are used to show the
efficiency of the presented method. Linear support vector machine (SVM) is
utilized to classify feature sets extracted with the presented method. The
obtained results show that the features extracted in this method have excellent
ability to classify faults without any additional feature selection; it is also
shown that EMD can improve or degrade features according to the utilized
feature reduction method.
|
Fast Methods for Recovering Sparse Parameters in Linear Low Rank Models | In this paper, we investigate the recovery of a sparse weight vector
(parameters vector) from a set of noisy linear combinations. However, only
partial information about the matrix representing the linear combinations is
available. Assuming a low-rank structure for the matrix, one natural solution
would be to first apply a matrix completion on the data, and then to solve the
resulting compressed sensing problem. In big data applications such as massive
MIMO and medical data, the matrix completion step imposes a huge computational
burden. Here, we propose to reduce the computational cost of the completion
task by ignoring the columns corresponding to zero elements in the sparse
vector. To this end, we employ a technique to initially approximate the support
of the sparse vector. We further propose to unify the partial matrix completion
and sparse vector recovery into an augmented four-step problem. Simulation
results reveal that the augmented approach achieves the best performance, while
both proposed methods outperform the natural two-step technique with
substantially less computational requirements.
|
Training LDCRF model on unsegmented sequences using Connectionist
Temporal Classification | Many machine learning problems such as speech recognition, gesture
recognition, and handwriting recognition are concerned with simultaneous
segmentation and labeling of sequence data. Latent-dynamic conditional random
field (LDCRF) is a well-known discriminative method that has been successfully
used for this task. However, LDCRF can only be trained with pre-segmented data
sequences in which the label of each frame is available apriori. In the realm
of neural networks, the invention of connectionist temporal classification
(CTC) made it possible to train recurrent neural networks on unsegmented
sequences with great success. In this paper, we use CTC to train an LDCRF model
on unsegmented sequences. Experimental results on two gesture recognition tasks
show that the proposed method outperforms LDCRFs, hidden Markov models, and
conditional random fields.
|
Exact gradient updates in time independent of output size for the
spherical loss family | An important class of problems involves training deep neural networks with
sparse prediction targets of very high dimension D. These occur naturally in
e.g. neural language models or the learning of word-embeddings, often posed as
predicting the probability of next words among a vocabulary of size D (e.g.
200,000). Computing the equally large, but typically non-sparse D-dimensional
output vector from a last hidden layer of reasonable dimension d (e.g. 500)
incurs a prohibitive O(Dd) computational cost for each example, as does
updating the $D \times d$ output weight matrix and computing the gradient
needed for backpropagation to previous layers. While efficient handling of
large sparse network inputs is trivial, the case of large sparse targets is
not, and has thus so far been sidestepped with approximate alternatives such as
hierarchical softmax or sampling-based approximations during training. In this
work we develop an original algorithmic approach which, for a family of loss
functions that includes squared error and spherical softmax, can compute the
exact loss, gradient update for the output weights, and gradient for
backpropagation, all in $O(d^{2})$ per example instead of $O(Dd)$, remarkably
without ever computing the D-dimensional output. The proposed algorithm yields
a speedup of up to $D/4d$ i.e. two orders of magnitude for typical sizes, for
that critical part of the computations that often dominates the training time
in this kind of network architecture.
|
Improved Recurrent Neural Networks for Session-based Recommendations | Recurrent neural networks (RNNs) were recently proposed for the session-based
recommendation task. The models showed promising improvements over traditional
recommendation approaches. In this work, we further study RNN-based models for
session-based recommendations. We propose the application of two techniques to
improve model performance, namely, data augmentation, and a method to account
for shifts in the input data distribution. We also empirically study the use of
generalised distillation, and a novel alternative model that directly predicts
item embeddings. Experiments on the RecSys Challenge 2015 dataset demonstrate
relative improvements of 12.8% and 14.8% over previously reported results on
the Recall@20 and Mean Reciprocal Rank@20 metrics respectively.
|
Supervised learning based on temporal coding in spiking neural networks | Gradient descent training techniques are remarkably successful in training
analog-valued artificial neural networks (ANNs). Such training techniques,
however, do not transfer easily to spiking networks due to the spike generation
hard non-linearity and the discrete nature of spike communication. We show that
in a feedforward spiking network that uses a temporal coding scheme where
information is encoded in spike times instead of spike rates, the network
input-output relation is differentiable almost everywhere. Moreover, this
relation is piece-wise linear after a transformation of variables. Methods for
training ANNs thus carry directly to the training of such spiking networks as
we show when training on the permutation invariant MNIST task. In contrast to
rate-based spiking networks that are often used to approximate the behavior of
ANNs, the networks we present spike much more sparsely and their behavior can
not be directly approximated by conventional ANNs. Our results highlight a new
approach for controlling the behavior of spiking networks with realistic
temporal dynamics, opening up the potential for using these networks to process
spike patterns with complex temporal information.
|
Out-of-Sample Extension for Dimensionality Reduction of Noisy Time
Series | This paper proposes an out-of-sample extension framework for a global
manifold learning algorithm (Isomap) that uses temporal information in
out-of-sample points in order to make the embedding more robust to noise and
artifacts. Given a set of noise-free training data and its embedding, the
proposed framework extends the embedding for a noisy time series. This is
achieved by adding a spatio-temporal compactness term to the optimization
objective of the embedding. To the best of our knowledge, this is the first
method for out-of-sample extension of manifold embeddings that leverages timing
information available for the extension set. Experimental results demonstrate
that our out-of-sample extension algorithm renders a more robust and accurate
embedding of sequentially ordered image data in the presence of various noise
and artifacts when compared to other timing-aware embeddings. Additionally, we
show that an out-of-sample extension framework based on the proposed algorithm
outperforms the state of the art in eye-gaze estimation.
|
Lifted Rule Injection for Relation Embeddings | Methods based on representation learning currently hold the state-of-the-art
in many natural language processing and knowledge base inference tasks. Yet, a
major challenge is how to efficiently incorporate commonsense knowledge into
such models. A recent approach regularizes relation and entity representations
by propositionalization of first-order logic rules. However,
propositionalization does not scale beyond domains with only few entities and
rules. In this paper we present a highly efficient method for incorporating
implication rules into distributed representations for automated knowledge base
construction. We map entity-tuple embeddings into an approximately Boolean
space and encourage a partial ordering over relation embeddings based on
implication rules mined from WordNet. Surprisingly, we find that the strong
restriction of the entity-tuple embedding space does not hurt the
expressiveness of the model and even acts as a regularizer that improves
generalization. By incorporating few commonsense rules, we achieve an increase
of 2 percentage points mean average precision over a matrix factorization
baseline, while observing a negligible increase in runtime.
|
A Reduction for Optimizing Lattice Submodular Functions with Diminishing
Returns | A function $f: \mathbb{Z}_+^E \rightarrow \mathbb{R}_+$ is DR-submodular if
it satisfies $f({\bf x} + \chi_i) -f ({\bf x}) \ge f({\bf y} + \chi_i) - f({\bf
y})$ for all ${\bf x}\le {\bf y}, i\in E$. Recently, the problem of maximizing
a DR-submodular function $f: \mathbb{Z}_+^E \rightarrow \mathbb{R}_+$ subject
to a budget constraint $\|{\bf x}\|_1 \leq B$ as well as additional constraints
has received significant attention \cite{SKIK14,SY15,MYK15,SY16}.
In this note, we give a generic reduction from the DR-submodular setting to
the submodular setting. The running time of the reduction and the size of the
resulting submodular instance depends only \emph{logarithmically} on $B$. Using
this reduction, one can translate the results for unconstrained and constrained
submodular maximization to the DR-submodular setting for many types of
constraints in a unified manner.
|
Gaussian Error Linear Units (GELUs) | We propose the Gaussian Error Linear Unit (GELU), a high-performing neural
network activation function. The GELU activation function is $x\Phi(x)$, where
$\Phi(x)$ the standard Gaussian cumulative distribution function. The GELU
nonlinearity weights inputs by their value, rather than gates inputs by their
sign as in ReLUs ($x\mathbf{1}_{x>0}$). We perform an empirical evaluation of
the GELU nonlinearity against the ReLU and ELU activations and find performance
improvements across all considered computer vision, natural language
processing, and speech tasks.
|
Symmetric and antisymmetric properties of solutions to kernel-based
machine learning problems | A particularly interesting instance of supervised learning with kernels is
when each training example is associated with two objects, as in pairwise
classification (Brunner et al., 2012), and in supervised learning of preference
relations (Herbrich et al., 1998). In these cases, one may want to embed
additional prior knowledge into the optimization problem associated with the
training of the learning machine, modeled, respectively, by the symmetry of its
optimal solution with respect to an exchange of order between the two objects,
and by its antisymmetry. Extending the approach proposed in (Brunner et al.,
2012) (where the only symmetric case was considered), we show, focusing on
support vector binary classification, how such embedding is possible through
the choice of a suitable pairwise kernel, which takes as inputs the individual
feature vectors and also the group feature vectors associated with the two
objects. We also prove that the symmetry/antisymmetry constraints still hold
when considering the sequence of suboptimal solutions generated by one version
of the Sequential Minimal Optimization (SMO) algorithm, and we present
numerical results supporting the theoretical findings. We conclude discussing
extensions of the main results to support vector regression, to transductive
support vector machines, and to several kinds of graph kernels, including
diffusion kernels.
|
A Learning Algorithm for Relational Logistic Regression: Preliminary
Results | Relational logistic regression (RLR) is a representation of conditional
probability in terms of weighted formulae for modelling multi-relational data.
In this paper, we develop a learning algorithm for RLR models. Learning an RLR
model from data consists of two steps: 1- learning the set of formulae to be
used in the model (a.k.a. structure learning) and learning the weight of each
formula (a.k.a. parameter learning). For structure learning, we deploy Schmidt
and Murphy's hierarchical assumption: first we learn a model with simple
formulae, then more complex formulae are added iteratively only if all their
sub-formulae have proven effective in previous learned models. For parameter
learning, we convert the problem into a non-relational learning problem and use
an off-the-shelf logistic regression learning algorithm from Weka, an
open-source machine learning tool, to learn the weights. We also indicate how
hidden features about the individuals can be incorporated into RLR to boost the
learning performance. We compare our learning algorithm to other structure and
parameter learning algorithms in the literature, and compare the performance of
RLR models to standard logistic regression and RDN-Boost on a modified version
of the MovieLens data-set.
|
A Local Density-Based Approach for Local Outlier Detection | This paper presents a simple but effective density-based outlier detection
approach with the local kernel density estimation (KDE). A Relative
Density-based Outlier Score (RDOS) is introduced to measure the local
outlierness of objects, in which the density distribution at the location of an
object is estimated with a local KDE method based on extended nearest neighbors
of the object. Instead of using only $k$ nearest neighbors, we further consider
reverse nearest neighbors and shared nearest neighbors of an object for density
distribution estimation. Some theoretical properties of the proposed RDOS
including its expected value and false alarm probability are derived. A
comprehensive experimental study on both synthetic and real-life data sets
demonstrates that our approach is more effective than state-of-the-art outlier
detection methods.
|
Estimating the class prior and posterior from noisy positives and
unlabeled data | We develop a classification algorithm for estimating posterior distributions
from positive-unlabeled data, that is robust to noise in the positive labels
and effective for high-dimensional data. In recent years, several algorithms
have been proposed to learn from positive-unlabeled data; however, many of
these contributions remain theoretical, performing poorly on real
high-dimensional data that is typically contaminated with noise. We build on
this previous work to develop two practical classification algorithms that
explicitly model the noise in the positive labels and utilize univariate
transforms built on discriminative classifiers. We prove that these univariate
transforms preserve the class prior, enabling estimation in the univariate
space and avoiding kernel density estimation for high-dimensional data. The
theoretical development and both parametric and nonparametric algorithms
proposed here constitutes an important step towards wide-spread use of robust
classification algorithms for positive-unlabeled data.
|
Alternating Back-Propagation for Generator Network | This paper proposes an alternating back-propagation algorithm for learning
the generator network model. The model is a non-linear generalization of factor
analysis. In this model, the mapping from the continuous latent factors to the
observed signal is parametrized by a convolutional neural network. The
alternating back-propagation algorithm iterates the following two steps: (1)
Inferential back-propagation, which infers the latent factors by Langevin
dynamics or gradient descent. (2) Learning back-propagation, which updates the
parameters given the inferred latent factors by gradient descent. The gradient
computations in both steps are powered by back-propagation, and they share most
of their code in common. We show that the alternating back-propagation
algorithm can learn realistic generator models of natural images, video
sequences, and sounds. Moreover, it can also be used to learn from incomplete
or indirect training data.
|
Clustering-Based Relational Unsupervised Representation Learning with an
Explicit Distributed Representation | The goal of unsupervised representation learning is to extract a new
representation of data, such that solving many different tasks becomes easier.
Existing methods typically focus on vectorized data and offer little support
for relational data, which additionally describe relationships among instances.
In this work we introduce an approach for relational unsupervised
representation learning. Viewing a relational dataset as a hypergraph, new
features are obtained by clustering vertices and hyperedges. To find a
representation suited for many relational learning tasks, a wide range of
similarities between relational objects is considered, e.g. feature and
structural similarities. We experimentally evaluate the proposed approach and
show that models learned on such latent representations perform better, have
lower complexity, and outperform the existing approaches on classification
tasks.
|
Theory reconstruction: a representation learning view on predicate
invention | With this positional paper we present a representation learning view on
predicate invention. The intention of this proposal is to bridge the relational
and deep learning communities on the problem of predicate invention. We propose
a theory reconstruction approach, a formalism that extends autoencoder approach
to representation learning to the relational settings. Our intention is to
start a discussion to define a unifying framework for predicate invention and
theory revision.
|
Reviving Threshold-Moving: a Simple Plug-in Bagging Ensemble for Binary
and Multiclass Imbalanced Data | Class imbalance presents a major hurdle in the application of data mining
methods. A common practice to deal with it is to create ensembles of
classifiers that learn from resampled balanced data. For example, bagged
decision trees combined with random undersampling (RUS) or the synthetic
minority oversampling technique (SMOTE). However, most of the resampling
methods entail asymmetric changes to the examples of different classes, which
in turn can introduce its own biases in the model. Furthermore, those methods
require a performance measure to be specified a priori before learning. An
alternative is to use a so-called threshold-moving method that a posteriori
changes the decision threshold of a model to counteract the imbalance, thus has
a potential to adapt to the performance measure of interest. Surprisingly,
little attention has been paid to the potential of combining bagging ensemble
with threshold-moving. In this paper, we present probability thresholding
bagging (PT-bagging), a versatile plug-in method that fills this gap. Contrary
to usual rebalancing practice, our method preserves the natural class
distribution of the data resulting in well calibrated posterior probabilities.
We also extend the proposed method to handle multiclass data. The method is
validated on binary and multiclass benchmark data sets. We perform analyses
that provide insights into the proposed method.
|
"Show me the cup": Reference with Continuous Representations | One of the most basic functions of language is to refer to objects in a
shared scene. Modeling reference with continuous representations is challenging
because it requires individuation, i.e., tracking and distinguishing an
arbitrary number of referents. We introduce a neural network model that, given
a definite description and a set of objects represented by natural images,
points to the intended object if the expression has a unique referent, or
indicates a failure, if it does not. The model, directly trained on reference
acts, is competitive with a pipeline manually engineered to perform the same
task, both when referents are purely visual, and when they are characterized by
a combination of visual and linguistic properties.
|
Adaptive Training of Random Mapping for Data Quantization | Data quantization learns encoding results of data with certain requirements,
and provides a broad perspective of many real-world applications to data
handling. Nevertheless, the results of encoder is usually limited to
multivariate inputs with the random mapping, and side information of binary
codes are hardly to mostly depict the original data patterns as possible. In
the literature, cosine based random quantization has attracted much attentions
due to its intrinsic bounded results. Nevertheless, it usually suffers from the
uncertain outputs, and information of original data fails to be fully preserved
in the reduced codes. In this work, a novel binary embedding method, termed
adaptive training quantization (ATQ), is proposed to learn the ideal transform
of random encoder, where the limitation of cosine random mapping is tackled. As
an adaptive learning idea, the reduced mapping is adaptively calculated with
idea of data group, while the bias of random transform is to be improved to
hold most matching information. Experimental results show that the proposed
method is able to obtain outstanding performance compared with other random
quantization methods.
|
European Union regulations on algorithmic decision-making and a "right
to explanation" | We summarize the potential impact that the European Union's new General Data
Protection Regulation will have on the routine use of machine learning
algorithms. Slated to take effect as law across the EU in 2018, it will
restrict automated individual decision-making (that is, algorithms that make
decisions based on user-level predictors) which "significantly affect" users.
The law will also effectively create a "right to explanation," whereby a user
can ask for an explanation of an algorithmic decision that was made about them.
We argue that while this law will pose large challenges for industry, it
highlights opportunities for computer scientists to take the lead in designing
algorithms and evaluation frameworks which avoid discrimination and enable
explanation.
|
Multi-View Kernel Consensus For Data Analysis | The input data features set for many data driven tasks is high-dimensional
while the intrinsic dimension of the data is low. Data analysis methods aim to
uncover the underlying low dimensional structure imposed by the low dimensional
hidden parameters by utilizing distance metrics that consider the set of
attributes as a single monolithic set. However, the transformation of the low
dimensional phenomena into the measured high dimensional observations might
distort the distance metric, This distortion can effect the desired estimated
low dimensional geometric structure. In this paper, we suggest to utilize the
redundancy in the attribute domain by partitioning the attributes into multiple
subsets we call views. The proposed methods utilize the agreement also called
consensus between different views to extract valuable geometric information
that unifies multiple views about the intrinsic relationships among several
different observations. This unification enhances the information that a single
view or a simple concatenations of views provides.
|
Active Ranking from Pairwise Comparisons and when Parametric Assumptions
Don't Help | We consider sequential or active ranking of a set of n items based on noisy
pairwise comparisons. Items are ranked according to the probability that a
given item beats a randomly chosen item, and ranking refers to partitioning the
items into sets of pre-specified sizes according to their scores. This notion
of ranking includes as special cases the identification of the top-k items and
the total ordering of the items. We first analyze a sequential ranking
algorithm that counts the number of comparisons won, and uses these counts to
decide whether to stop, or to compare another pair of items, chosen based on
confidence intervals specified by the data collected up to that point. We prove
that this algorithm succeeds in recovering the ranking using a number of
comparisons that is optimal up to logarithmic factors. This guarantee does not
require any structural properties of the underlying pairwise probability
matrix, unlike a significant body of past work on pairwise ranking based on
parametric models such as the Thurstone or Bradley-Terry-Luce models. It has
been a long-standing open question as to whether or not imposing these
parametric assumptions allows for improved ranking algorithms. For stochastic
comparison models, in which the pairwise probabilities are bounded away from
zero, our second contribution is to resolve this issue by proving a lower bound
for parametric models. This shows, perhaps surprisingly, that these popular
parametric modeling choices offer at most logarithmic gains for stochastic
comparisons.
|
Technical Report: Towards a Universal Code Formatter through Machine
Learning | There are many declarative frameworks that allow us to implement code
formatters relatively easily for any specific language, but constructing them
is cumbersome. The first problem is that "everybody" wants to format their code
differently, leading to either many formatter variants or a ridiculous number
of configuration options. Second, the size of each implementation scales with a
language's grammar size, leading to hundreds of rules.
In this paper, we solve the formatter construction problem using a novel
approach, one that automatically derives formatters for any given language
without intervention from a language expert. We introduce a code formatter
called CodeBuff that uses machine learning to abstract formatting rules from a
representative corpus, using a carefully designed feature set. Our experiments
on Java, SQL, and ANTLR grammars show that CodeBuff is efficient, has excellent
accuracy, and is grammar invariant for a given language. It also generalizes to
a 4th language tested during manuscript preparation.
|
Defending Non-Bayesian Learning against Adversarial Attacks | This paper addresses the problem of non-Bayesian learning over multi-agent
networks, where agents repeatedly collect partially informative observations
about an unknown state of the world, and try to collaboratively learn the true
state. We focus on the impact of the adversarial agents on the performance of
consensus-based non-Bayesian learning, where non-faulty agents combine local
learning updates with consensus primitives. In particular, we consider the
scenario where an unknown subset of agents suffer Byzantine faults -- agents
suffering Byzantine faults behave arbitrarily. Two different learning rules are
proposed.
|
Exact Lower Bounds for the Agnostic Probably-Approximately-Correct (PAC)
Machine Learning Model | We provide an exact non-asymptotic lower bound on the minimax expected excess
risk (EER) in the agnostic probably-ap\-proximately-correct (PAC) machine
learning classification model and identify minimax learning algorithms as
certain maximally symmetric and minimally randomized "voting" procedures. Based
on this result, an exact asymptotic lower bound on the minimax EER is provided.
This bound is of the simple form $c_\infty/\sqrt{\nu}$ as $\nu\to\infty$, where
$c_\infty=0.16997\dots$ is a universal constant, $\nu=m/d$, $m$ is the size of
the training sample, and $d$ is the Vapnik--Chervonenkis dimension of the
hypothesis class. It is shown that the differences between these asymptotic and
non-asymptotic bounds, as well as the differences between these two bounds and
the maximum EER of any learning algorithms that minimize the empirical risk,
are asymptotically negligible, and all these differences are due to ties in the
mentioned "voting" procedures. A few easy to compute non-asymptotic lower
bounds on the minimax EER are also obtained, which are shown to be close to the
exact asymptotic lower bound $c_\infty/\sqrt{\nu}$ even for rather small values
of the ratio $\nu=m/d$. As an application of these results, we substantially
improve existing lower bounds on the tail probability of the excess risk. Among
the tools used are Bayes estimation and apparently new identities and
inequalities for binomial distributions.
|
subgraph2vec: Learning Distributed Representations of Rooted Sub-graphs
from Large Graphs | In this paper, we present subgraph2vec, a novel approach for learning latent
representations of rooted subgraphs from large graphs inspired by recent
advancements in Deep Learning and Graph Kernels. These latent representations
encode semantic substructure dependencies in a continuous vector space, which
is easily exploited by statistical models for tasks such as graph
classification, clustering, link prediction and community detection.
subgraph2vec leverages on local information obtained from neighbourhoods of
nodes to learn their latent representations in an unsupervised fashion. We
demonstrate that subgraph vectors learnt by our approach could be used in
conjunction with classifiers such as CNNs, SVMs and relational data clustering
algorithms to achieve significantly superior accuracies. Also, we show that the
subgraph vectors could be used for building a deep learning variant of
Weisfeiler-Lehman graph kernel. Our experiments on several benchmark and
large-scale real-world datasets reveal that subgraph2vec achieves significant
improvements in accuracies over existing graph kernels on both supervised and
unsupervised learning tasks. Specifically, on two realworld program analysis
tasks, namely, code clone and malware detection, subgraph2vec outperforms
state-of-the-art kernels by more than 17% and 4%, respectively.
|
Non-linear Label Ranking for Large-scale Prediction of Long-Term User
Interests | We consider the problem of personalization of online services from the
viewpoint of ad targeting, where we seek to find the best ad categories to be
shown to each user, resulting in improved user experience and increased
advertisers' revenue. We propose to address this problem as a task of ranking
the ad categories depending on a user's preference, and introduce a novel label
ranking approach capable of efficiently learning non-linear, highly accurate
models in large-scale settings. Experiments on a real-world advertising data
set with more than 3.2 million users show that the proposed algorithm
outperforms the existing solutions in terms of both rank loss and top-K
retrieval performance, strongly suggesting the benefit of using the proposed
model on large-scale ranking problems.
|
Decision making via semi-supervised machine learning techniques | Semi-supervised learning (SSL) is a class of supervised learning tasks and
techniques that also exploits the unlabeled data for training. SSL
significantly reduces labeling related costs and is able to handle large data
sets. The primary objective is the extraction of robust inference rules.
Decision support systems (DSSs) who utilize SSL have significant advantages.
Only a small amount of labelled data is required for the initialization. Then,
new (unlabeled) data can be utilized and improve system's performance. Thus,
the DSS is continuously adopted to new conditions, with minimum effort.
Techniques which are cost effective and easily adopted to dynamic systems, can
be beneficial for many practical applications. Such applications fields are:
(a) industrial assembly lines monitoring, (b) sea border surveillance, (c)
elders' falls detection, (d) transportation tunnels inspection, (e) concrete
foundation piles defect recognition, (f) commercial sector companies financial
assessment and (g) image advanced filtering for cultural heritage applications.
|
A Distributional Semantics Approach to Implicit Language Learning | In the present paper we show that distributional information is particularly
important when considering concept availability under implicit language
learning conditions. Based on results from different behavioural experiments we
argue that the implicit learnability of semantic regularities depends on the
degree to which the relevant concept is reflected in language use. In our
simulations, we train a Vector-Space model on either an English or a Chinese
corpus and then feed the resulting representations to a feed-forward neural
network. The task of the neural network was to find a mapping between the word
representations and the novel words. Using datasets from four behavioural
experiments, which used different semantic manipulations, we were able to
obtain learning patterns very similar to those obtained by humans.
|
Actor-critic versus direct policy search: a comparison based on sample
complexity | Sample efficiency is a critical property when optimizing policy parameters
for the controller of a robot. In this paper, we evaluate two state-of-the-art
policy optimization algorithms. One is a recent deep reinforcement learning
method based on an actor-critic algorithm, Deep Deterministic Policy Gradient
(DDPG), that has been shown to perform well on various control benchmarks. The
other one is a direct policy search method, Covariance Matrix Adaptation
Evolution Strategy (CMA-ES), a black-box optimization method that is widely
used for robot learning. The algorithms are evaluated on a continuous version
of the mountain car benchmark problem, so as to compare their sample
complexity. From a preliminary analysis, we expect DDPG to be more sample
efficient than CMA-ES, which is confirmed by our experimental results.
|
Disease Trajectory Maps | Medical researchers are coming to appreciate that many diseases are in fact
complex, heterogeneous syndromes composed of subpopulations that express
different variants of a related complication. Time series data extracted from
individual electronic health records (EHR) offer an exciting new way to study
subtle differences in the way these diseases progress over time. In this paper,
we focus on answering two questions that can be asked using these databases of
time series. First, we want to understand whether there are individuals with
similar disease trajectories and whether there are a small number of degrees of
freedom that account for differences in trajectories across the population.
Second, we want to understand how important clinical outcomes are associated
with disease trajectories. To answer these questions, we propose the Disease
Trajectory Map (DTM), a novel probabilistic model that learns low-dimensional
representations of sparse and irregularly sampled time series. We propose a
stochastic variational inference algorithm for learning the DTM that allows the
model to scale to large modern medical datasets. To demonstrate the DTM, we
analyze data collected on patients with the complex autoimmune disease,
scleroderma. We find that DTM learns meaningful representations of disease
trajectories and that the representations are significantly associated with
important clinical outcomes.
|
A Semi-Definite Programming approach to low dimensional embedding for
unsupervised clustering | This paper proposes a variant of the method of Gu\'edon and Verhynin for
estimating the cluster matrix in the Mixture of Gaussians framework via
Semi-Definite Programming. A clustering oriented embedding is deduced from this
estimate. The procedure is suitable for very high dimensional data because it
is based on pairwise distances only. Theoretical garantees are provided and an
eigenvalue optimisation approach is proposed for computing the embedding. The
performance of the method is illustrated via Monte Carlo experiements and
comparisons with other embeddings from the literature.
|
Model-Free Trajectory-based Policy Optimization with Monotonic
Improvement | Many of the recent trajectory optimization algorithms alternate between
linear approximation of the system dynamics around the mean trajectory and
conservative policy update. One way of constraining the policy change is by
bounding the Kullback-Leibler (KL) divergence between successive policies.
These approaches already demonstrated great experimental success in challenging
problems such as end-to-end control of physical systems. However, the linear
approximation of the system dynamics can introduce a bias in the policy update
and prevent convergence to the optimal policy. In this article, we propose a
new model-free trajectory-based policy optimization algorithm with guaranteed
monotonic improvement. The algorithm backpropagates a local, quadratic and
time-dependent \qfunc~learned from trajectory data instead of a model of the
system dynamics. Our policy update ensures exact KL-constraint satisfaction
without simplifying assumptions on the system dynamics. We experimentally
demonstrate on highly non-linear control tasks the improvement in performance
of our algorithm in comparison to approaches linearizing the system dynamics.
In order to show the monotonic improvement of our algorithm, we additionally
conduct a theoretical analysis of our policy update scheme to derive a lower
bound of the change in policy return between successive iterations.
|
Tighter bounds lead to improved classifiers | The standard approach to supervised classification involves the minimization
of a log-loss as an upper bound to the classification error. While this is a
tight bound early on in the optimization, it overemphasizes the influence of
incorrectly classified examples far from the decision boundary. Updating the
upper bound during the optimization leads to improved classification rates
while transforming the learning into a sequence of minimization problems. In
addition, in the context where the classifier is part of a larger system, this
modification makes it possible to link the performance of the classifier to
that of the whole system, allowing the seamless introduction of external
constraints.
|
Learning Concept Taxonomies from Multi-modal Data | We study the problem of automatically building hypernym taxonomies from
textual and visual data. Previous works in taxonomy induction generally ignore
the increasingly prominent visual data, which encode important perceptual
semantics. Instead, we propose a probabilistic model for taxonomy induction by
jointly leveraging text and images. To avoid hand-crafted feature engineering,
we design end-to-end features based on distributed representations of images
and words. The model is discriminatively trained given a small set of existing
ontologies and is capable of building full taxonomies from scratch for a
collection of unseen conceptual label items with associated images. We evaluate
our model and features on the WordNet hierarchies, where our system outperforms
previous approaches by a large gap.
|
Learning without Forgetting | When building a unified vision system or gradually adding new capabilities to
a system, the usual assumption is that training data for all tasks is always
available. However, as the number of tasks grows, storing and retraining on
such data becomes infeasible. A new problem arises where we add new
capabilities to a Convolutional Neural Network (CNN), but the training data for
its existing capabilities are unavailable. We propose our Learning without
Forgetting method, which uses only new task data to train the network while
preserving the original capabilities. Our method performs favorably compared to
commonly used feature extraction and fine-tuning adaption techniques and
performs similarly to multitask learning that uses original task data we assume
unavailable. A more surprising observation is that Learning without Forgetting
may be able to replace fine-tuning with similar old and new task datasets for
improved new task performance.
|
Dimension-Free Iteration Complexity of Finite Sum Optimization Problems | Many canonical machine learning problems boil down to a convex optimization
problem with a finite sum structure. However, whereas much progress has been
made in developing faster algorithms for this setting, the inherent limitations
of these problems are not satisfactorily addressed by existing lower bounds.
Indeed, current bounds focus on first-order optimization algorithms, and only
apply in the often unrealistic regime where the number of iterations is less
than $\mathcal{O}(d/n)$ (where $d$ is the dimension and $n$ is the number of
samples). In this work, we extend the framework of (Arjevani et al., 2015) to
provide new lower bounds, which are dimension-free, and go beyond the
assumptions of current bounds, thereby covering standard finite sum
optimization methods, e.g., SAG, SAGA, SVRG, SDCA without duality, as well as
stochastic coordinate-descent methods, such as SDCA and accelerated proximal
SDCA.
|
Convolutional Neural Networks on Graphs with Fast Localized Spectral
Filtering | In this work, we are interested in generalizing convolutional neural networks
(CNNs) from low-dimensional regular grids, where image, video and speech are
represented, to high-dimensional irregular domains, such as social networks,
brain connectomes or words' embedding, represented by graphs. We present a
formulation of CNNs in the context of spectral graph theory, which provides the
necessary mathematical background and efficient numerical schemes to design
fast localized convolutional filters on graphs. Importantly, the proposed
technique offers the same linear computational complexity and constant learning
complexity as classical CNNs, while being universal to any graph structure.
Experiments on MNIST and 20NEWS demonstrate the ability of this novel deep
learning system to learn local, stationary, and compositional features on
graphs.
|
On Approximate Dynamic Programming with Multivariate Splines for
Adaptive Control | We define a SDP framework based on the RLSTD algorithm and multivariate
simplex B-splines. We introduce a local forget factor capable of preserving the
continuity of the simplex splines. This local forget factor is integrated with
the RLSTD algorithm, resulting in a modified RLSTD algorithm that is capable of
tracking time-varying systems. We present the results of two numerical
experiments, one validating SDP and comparing it with NDP and another to show
the advantages of the modified RLSTD algorithm over the original. While SDP
requires more computations per time-step, the experiment shows that for the
same amount of function approximator parameters, there is an increase in
performance in terms of stability and learning rate compared to NDP. The second
experiment shows that SDP in combination with the modified RLSTD algorithm
allows for faster recovery compared to the original RLSTD algorithm when system
parameters are altered, paving the way for an adaptive high-performance
non-linear control method.
|
Asymptotically Optimal Algorithms for Budgeted Multiple Play Bandits | We study a generalization of the multi-armed bandit problem with multiple
plays where there is a cost associated with pulling each arm and the agent has
a budget at each time that dictates how much she can expect to spend. We derive
an asymptotic regret lower bound for any uniformly efficient algorithm in our
setting. We then study a variant of Thompson sampling for Bernoulli rewards and
a variant of KL-UCB for both single-parameter exponential families and bounded,
finitely supported rewards. We show these algorithms are asymptotically
optimal, both in rateand leading problem-dependent constants, including in the
thick margin setting where multiple arms fall on the decision boundary.
|
Vote-boosting ensembles | Vote-boosting is a sequential ensemble learning method in which the
individual classifiers are built on different weighted versions of the training
data. To build a new classifier, the weight of each training instance is
determined in terms of the degree of disagreement among the current ensemble
predictions for that instance. For low class-label noise levels, especially
when simple base learners are used, emphasis should be made on instances for
which the disagreement rate is high. When more flexible classifiers are used
and as the noise level increases, the emphasis on these uncertain instances
should be reduced. In fact, at sufficiently high levels of class-label noise,
the focus should be on instances on which the ensemble classifiers agree. The
optimal type of emphasis can be automatically determined using
cross-validation. An extensive empirical analysis using the beta distribution
as emphasis function illustrates that vote-boosting is an effective method to
generate ensembles that are both accurate and robust.
|
A Model Explanation System: Latest Updates and Extensions | We propose a general model explanation system (MES) for "explaining" the
output of black box classifiers. This paper describes extensions to Turner
(2015), which is referred to frequently in the text. We use the motivating
example of a classifier trained to detect fraud in a credit card transaction
history. The key aspect is that we provide explanations applicable to a single
prediction, rather than provide an interpretable set of parameters. We focus on
explaining positive predictions (alerts). However, the presented methodology is
symmetrically applicable to negative predictions.
|
Performance Based Evaluation of Various Machine Learning Classification
Techniques for Chronic Kidney Disease Diagnosis | Areas where Artificial Intelligence (AI) & related fields are finding their
applications are increasing day by day, moving from core areas of computer
science they are finding their applications in various other domains.In recent
times Machine Learning i.e. a sub-domain of AI has been widely used in order to
assist medical experts and doctors in the prediction, diagnosis and prognosis
of various diseases and other medical disorders. In this manuscript the authors
applied various machine learning algorithms to a problem in the domain of
medical diagnosis and analyzed their efficiency in predicting the results. The
problem selected for the study is the diagnosis of the Chronic Kidney
Disease.The dataset used for the study consists of 400 instances and 24
attributes. The authors evaluated 12 classification techniques by applying them
to the Chronic Kidney Disease data. In order to calculate efficiency, results
of the prediction by candidate methods were compared with the actual medical
results of the subject.The various metrics used for performance evaluation are
predictive accuracy, precision, sensitivity and specificity. The results
indicate that decision-tree performed best with nearly the accuracy of 98.6%,
sensitivity of 0.9720, precision of 1 and specificity of 1.
|
A Permutation-based Model for Crowd Labeling: Optimal Estimation and
Robustness | The task of aggregating and denoising crowd-labeled data has gained increased
significance with the advent of crowdsourcing platforms and massive datasets.
We propose a permutation-based model for crowd labeled data that is a
significant generalization of the classical Dawid-Skene model, and introduce a
new error metric by which to compare different estimators. We derive global
minimax rates for the permutation-based model that are sharp up to logarithmic
factors, and match the minimax lower bounds derived under the simpler
Dawid-Skene model. We then design two computationally-efficient estimators: the
WAN estimator for the setting where the ordering of workers in terms of their
abilities is approximately known, and the OBI-WAN estimator where that is not
known. For each of these estimators, we provide non-asymptotic bounds on their
performance. We conduct synthetic simulations and experiments on real-world
crowdsourcing data, and the experimental results corroborate our theoretical
findings.
|
Review Based Rating Prediction | Recommendation systems are an important units in today's e-commerce
applications, such as targeted advertising, personalized marketing and
information retrieval. In recent years, the importance of contextual
information has motivated generation of personalized recommendations according
to the available contextual information of users.
Compared to the traditional systems which mainly utilize users' rating
history, review-based recommendation hopefully provide more relevant results to
users. We introduce a review-based recommendation approach that obtains
contextual information by mining user reviews. The proposed approach relate to
features obtained by analyzing textual reviews using methods developed in
Natural Language Processing (NLP) and information retrieval discipline to
compute a utility function over a given item. An item utility is a measure that
shows how much it is preferred according to user's current context.
In our system, the context inference is modeled as similarity between the
users reviews history and the item reviews history. As an example application,
we used our method to mine contextual data from customers' reviews of movies
and use it to produce review-based rating prediction. The predicted ratings can
generate recommendations that are item-based and should appear at the
recommended items list in the product page. Our evaluations suggest that our
system can help produce better prediction rating scores in comparison to the
standard prediction methods.
|
Ballpark Learning: Estimating Labels from Rough Group Comparisons | We are interested in estimating individual labels given only coarse,
aggregated signal over the data points. In our setting, we receive sets
("bags") of unlabeled instances with constraints on label proportions. We relax
the unrealistic assumption of known label proportions, made in previous work;
instead, we assume only to have upper and lower bounds, and constraints on bag
differences. We motivate the problem, propose an intuitive formulation and
algorithm, and apply our methods to real-world scenarios. Across several
domains, we show how using only proportion constraints and no labeled examples,
we can achieve surprisingly high accuracy. In particular, we demonstrate how to
predict income level using rough stereotypes and how to perform sentiment
analysis using very little information. We also apply our method to guide
exploratory analysis, recovering geographical differences in twitter dialect.
|
Dynamic Neural Turing Machine with Soft and Hard Addressing Schemes | We extend neural Turing machine (NTM) model into a dynamic neural Turing
machine (D-NTM) by introducing a trainable memory addressing scheme. This
addressing scheme maintains for each memory cell two separate vectors, content
and address vectors. This allows the D-NTM to learn a wide variety of
location-based addressing strategies including both linear and nonlinear ones.
We implement the D-NTM with both continuous, differentiable and discrete,
non-differentiable read/write mechanisms. We investigate the mechanisms and
effects of learning to read and write into a memory through experiments on
Facebook bAbI tasks using both a feedforward and GRUcontroller. The D-NTM is
evaluated on a set of Facebook bAbI tasks and shown to outperform NTM and LSTM
baselines. We have done extensive analysis of our model and different
variations of NTM on bAbI task. We also provide further experimental results on
sequential pMNIST, Stanford Natural Language Inference, associative recall and
copy tasks.
|
Unsupervised Learning with Imbalanced Data via Structure Consolidation
Latent Variable Model | Unsupervised learning on imbalanced data is challenging because, when given
imbalanced data, current model is often dominated by the major category and
ignores the categories with small amount of data. We develop a latent variable
model that can cope with imbalanced data by dividing the latent space into a
shared space and a private space. Based on Gaussian Process Latent Variable
Models, we propose a new kernel formulation that enables the separation of
latent space and derives an efficient variational inference method. The
performance of our model is demonstrated with an imbalanced medical image
dataset.
|
Multi-class classification: mirror descent approach | We consider the problem of multi-class classification and a stochastic opti-
mization approach to it. We derive risk bounds for stochastic mirror descent
algorithm and provide examples of set geometries that make the use of the
algorithm efficient in terms of error in k.
|
Fractal Dimension Pattern Based Multiresolution Analysis for Rough
Estimator of Person-Dependent Audio Emotion Recognition | As a general means of expression, audio analysis and recognition has
attracted much attentions for its wide applications in real-life world. Audio
emotion recognition (AER) attempts to understand emotional states of human with
the given utterance signals, and has been studied abroad for its further
development on friendly human-machine interfaces. Distinguish from other
existing works, the person-dependent patterns of audio emotions are conducted,
and fractal dimension features are calculated for acoustic feature extraction.
Furthermore, it is able to efficiently learn intrinsic characteristics of
auditory emotions, while the utterance features are learned from fractal
dimensions of each sub-bands. Experimental results show the proposed method is
able to provide comparative performance for audio emotion recognition.
|
Randomized block proximal damped Newton method for composite
self-concordant minimization | In this paper we consider the composite self-concordant (CSC) minimization
problem, which minimizes the sum of a self-concordant function $f$ and a
(possibly nonsmooth) proper closed convex function $g$. The CSC minimization is
the cornerstone of the path-following interior point methods for solving a
broad class of convex optimization problems. It has also found numerous
applications in machine learning. The proximal damped Newton (PDN) methods have
been well studied in the literature for solving this problem that enjoy a nice
iteration complexity. Given that at each iteration these methods typically
require evaluating or accessing the Hessian of $f$ and also need to solve a
proximal Newton subproblem, the cost per iteration can be prohibitively high
when applied to large-scale problems. Inspired by the recent success of block
coordinate descent methods, we propose a randomized block proximal damped
Newton (RBPDN) method for solving the CSC minimization. Compared to the PDN
methods, the computational cost per iteration of RBPDN is usually significantly
lower. The computational experiment on a class of regularized logistic
regression problems demonstrate that RBPDN is indeed promising in solving
large-scale CSC minimization problems. The convergence of RBPDN is also
analyzed in the paper. In particular, we show that RBPDN is globally convergent
when $g$ is Lipschitz continuous. It is also shown that RBPDN enjoys a local
linear convergence. Moreover, we show that for a class of $g$ including the
case where $g$ is Lipschitz differentiable, RBPDN enjoys a global linear
convergence. As a striking consequence, it shows that the classical damped
Newton methods [22,40] and the PDN [31] for such $g$ are globally linearly
convergent, which was previously unknown in the literature. Moreover, this
result can be used to sharpen the existing iteration complexity of these
methods.
|
Combining Gradient Boosting Machines with Collective Inference to
Predict Continuous Values | Gradient boosting of regression trees is a competitive procedure for learning
predictive models of continuous data that fits the data with an additive
non-parametric model. The classic version of gradient boosting assumes that the
data is independent and identically distributed. However, relational data with
interdependent, linked instances is now common and the dependencies in such
data can be exploited to improve predictive performance. Collective inference
is one approach to exploit relational correlation patterns and significantly
reduce classification error. However, much of the work on collective learning
and inference has focused on discrete prediction tasks rather than continuous.
%target values has not got that attention in terms of collective inference. In
this work, we investigate how to combine these two paradigms together to
improve regression in relational domains. Specifically, we propose a boosting
algorithm for learning a collective inference model that predicts a continuous
target variable. In the algorithm, we learn a basic relational model,
collectively infer the target values, and then iteratively learn relational
models to predict the residuals. We evaluate our proposed algorithm on a real
network dataset and show that it outperforms alternative boosting methods.
However, our investigation also revealed that the relational features interact
together to produce better predictions.
|
Less-forgetting Learning in Deep Neural Networks | A catastrophic forgetting problem makes deep neural networks forget the
previously learned information, when learning data collected in new
environments, such as by different sensors or in different light conditions.
This paper presents a new method for alleviating the catastrophic forgetting
problem. Unlike previous research, our method does not use any information from
the source domain. Surprisingly, our method is very effective to forget less of
the information in the source domain, and we show the effectiveness of our
method using several experiments. Furthermore, we observed that the forgetting
problem occurs between mini-batches when performing general training processes
using stochastic gradient descent methods, and this problem is one of the
factors that degrades generalization performance of the network. We also try to
solve this problem using the proposed method. Finally, we show our
less-forgetting learning method is also helpful to improve the performance of
deep neural networks in terms of recognition rates.
|
Deep Learning with Differential Privacy | Machine learning techniques based on neural networks are achieving remarkable
results in a wide variety of domains. Often, the training of models requires
large, representative datasets, which may be crowdsourced and contain sensitive
information. The models should not expose private information in these
datasets. Addressing this goal, we develop new algorithmic techniques for
learning and a refined analysis of privacy costs within the framework of
differential privacy. Our implementation and experiments demonstrate that we
can train deep neural networks with non-convex objectives, under a modest
privacy budget, and at a manageable cost in software complexity, training
efficiency, and model quality.
|
Missing Data Estimation in High-Dimensional Datasets: A Swarm
Intelligence-Deep Neural Network Approach | In this paper, we examine the problem of missing data in high-dimensional
datasets by taking into consideration the Missing Completely at Random and
Missing at Random mechanisms, as well as theArbitrary missing pattern.
Additionally, this paper employs a methodology based on Deep Learning and Swarm
Intelligence algorithms in order to provide reliable estimates for missing
data. The deep learning technique is used to extract features from the input
data via an unsupervised learning approach by modeling the data distribution
based on the input. This deep learning technique is then used as part of the
objective function for the swarm intelligence technique in order to estimate
the missing data after a supervised fine-tuning phase by minimizing an error
function based on the interrelationship and correlation between features in the
dataset. The investigated methodology in this paper therefore has longer
running times, however, the promising potential outcomes justify the trade-off.
Also, basic knowledge of statistics is presumed.
|
Efficient and Consistent Robust Time Series Analysis | We study the problem of robust time series analysis under the standard
auto-regressive (AR) time series model in the presence of arbitrary outliers.
We devise an efficient hard thresholding based algorithm which can obtain a
consistent estimate of the optimal AR model despite a large fraction of the
time series points being corrupted. Our algorithm alternately estimates the
corrupted set of points and the model parameters, and is inspired by recent
advances in robust regression and hard-thresholding methods. However, a direct
application of existing techniques is hindered by a critical difference in the
time-series domain: each point is correlated with all previous points rendering
existing tools inapplicable directly. We show how to overcome this hurdle using
novel proof techniques. Using our techniques, we are also able to provide the
first efficient and provably consistent estimator for the robust regression
problem where a standard linear observation model with white additive noise is
corrupted arbitrarily. We illustrate our methods on synthetic datasets and show
that our methods indeed are able to consistently recover the optimal parameters
despite a large fraction of points being corrupted.
|
LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection | Mechanical devices such as engines, vehicles, aircrafts, etc., are typically
instrumented with numerous sensors to capture the behavior and health of the
machine. However, there are often external factors or variables which are not
captured by sensors leading to time-series which are inherently unpredictable.
For instance, manual controls and/or unmonitored environmental conditions or
load may lead to inherently unpredictable time-series. Detecting anomalies in
such scenarios becomes challenging using standard approaches based on
mathematical models that rely on stationarity, or prediction models that
utilize prediction errors to detect anomalies. We propose a Long Short Term
Memory Networks based Encoder-Decoder scheme for Anomaly Detection (EncDec-AD)
that learns to reconstruct 'normal' time-series behavior, and thereafter uses
reconstruction error to detect anomalies. We experiment with three publicly
available quasi predictable time-series datasets: power demand, space shuttle,
and ECG, and two real-world engine datasets with both predictive and
unpredictable behavior. We show that EncDec-AD is robust and can detect
anomalies from predictable, unpredictable, periodic, aperiodic, and
quasi-periodic time-series. Further, we show that EncDec-AD is able to detect
anomalies from short time-series (length as small as 30) as well as long
time-series (length as large as 500).
|
Why is Posterior Sampling Better than Optimism for Reinforcement
Learning? | Computational results demonstrate that posterior sampling for reinforcement
learning (PSRL) dramatically outperforms algorithms driven by optimism, such as
UCRL2. We provide insight into the extent of this performance boost and the
phenomenon that drives it. We leverage this insight to establish an
$\tilde{O}(H\sqrt{SAT})$ Bayesian expected regret bound for PSRL in
finite-horizon episodic Markov decision processes, where $H$ is the horizon,
$S$ is the number of states, $A$ is the number of actions and $T$ is the time
elapsed. This improves upon the best previous bound of $\tilde{O}(H S
\sqrt{AT})$ for any reinforcement learning algorithm.
|
Permutation Invariant Training of Deep Models for Speaker-Independent
Multi-talker Speech Separation | We propose a novel deep learning model, which supports permutation invariant
training (PIT), for speaker independent multi-talker speech separation,
commonly known as the cocktail-party problem. Different from most of the prior
arts that treat speech separation as a multi-class regression problem and the
deep clustering technique that considers it a segmentation (or clustering)
problem, our model optimizes for the separation regression error, ignoring the
order of mixing sources. This strategy cleverly solves the long-lasting label
permutation problem that has prevented progress on deep learning based
techniques for speech separation. Experiments on the equal-energy mixing setup
of a Danish corpus confirms the effectiveness of PIT. We believe improvements
built upon PIT can eventually solve the cocktail-party problem and enable
real-world adoption of, e.g., automatic meeting transcription and multi-party
human-computer interaction, where overlapping speech is common.
|
Convergence Rate of Frank-Wolfe for Non-Convex Objectives | We give a simple proof that the Frank-Wolfe algorithm obtains a stationary
point at a rate of $O(1/\sqrt{t})$ on non-convex objectives with a Lipschitz
continuous gradient. Our analysis is affine invariant and is the first, to the
best of our knowledge, giving a similar rate to what was already proven for
projected gradient methods (though on slightly different measures of
stationarity).
|
A scaled Bregman theorem with applications | Bregman divergences play a central role in the design and analysis of a range
of machine learning algorithms. This paper explores the use of Bregman
divergences to establish reductions between such algorithms and their analyses.
We present a new scaled isodistortion theorem involving Bregman divergences
(scaled Bregman theorem for short) which shows that certain "Bregman
distortions'" (employing a potentially non-convex generator) may be exactly
re-written as a scaled Bregman divergence computed over transformed data.
Admissible distortions include geodesic distances on curved manifolds and
projections or gauge-normalisation, while admissible data include scalars,
vectors and matrices.
Our theorem allows one to leverage to the wealth and convenience of Bregman
divergences when analysing algorithms relying on the aforementioned Bregman
distortions. We illustrate this with three novel applications of our theorem: a
reduction from multi-class density ratio to class-probability estimation, a new
adaptive projection free yet norm-enforcing dual norm mirror descent algorithm,
and a reduction from clustering on flat manifolds to clustering on curved
manifolds. Experiments on each of these domains validate the analyses and
suggest that the scaled Bregman theorem might be a worthy addition to the
popular handful of Bregman divergence properties that have been pervasive in
machine learning.
|
Domain Adaptation for Neural Networks by Parameter Augmentation | We propose a simple domain adaptation method for neural networks in a
supervised setting. Supervised domain adaptation is a way of improving the
generalization performance on the target domain by using the source domain
dataset, assuming that both of the datasets are labeled. Recently, recurrent
neural networks have been shown to be successful on a variety of NLP tasks such
as caption generation; however, the existing domain adaptation techniques are
limited to (1) tune the model parameters by the target dataset after the
training by the source dataset, or (2) design the network to have dual output,
one for the source domain and the other for the target domain. Reformulating
the idea of the domain adaptation technique proposed by Daume (2007), we
propose a simple domain adaptation method, which can be applied to neural
networks trained with a cross-entropy loss. On captioning datasets, we show
performance improvements over other domain adaptation methods.
|
Learning Relational Dependency Networks for Relation Extraction | We consider the task of KBP slot filling -- extracting relation information
from newswire documents for knowledge base construction. We present our
pipeline, which employs Relational Dependency Networks (RDNs) to learn
linguistic patterns for relation extraction. Additionally, we demonstrate how
several components such as weak supervision, word2vec features, joint learning
and the use of human advice, can be incorporated in this relational framework.
We evaluate the different components in the benchmark KBP 2015 task and show
that RDNs effectively model a diverse set of features and perform competitively
with current state-of-the-art relation extraction.
|
Decoding the Encoding of Functional Brain Networks: an fMRI
Classification Comparison of Non-negative Matrix Factorization (NMF),
Independent Component Analysis (ICA), and Sparse Coding Algorithms | Brain networks in fMRI are typically identified using spatial independent
component analysis (ICA), yet mathematical constraints such as sparse coding
and positivity both provide alternate biologically-plausible frameworks for
generating brain networks. Non-negative Matrix Factorization (NMF) would
suppress negative BOLD signal by enforcing positivity. Spatial sparse coding
algorithms ($L1$ Regularized Learning and K-SVD) would impose local
specialization and a discouragement of multitasking, where the total observed
activity in a single voxel originates from a restricted number of possible
brain networks.
The assumptions of independence, positivity, and sparsity to encode
task-related brain networks are compared; the resulting brain networks for
different constraints are used as basis functions to encode the observed
functional activity at a given time point. These encodings are decoded using
machine learning to compare both the algorithms and their assumptions, using
the time series weights to predict whether a subject is viewing a video,
listening to an audio cue, or at rest, in 304 fMRI scans from 51 subjects.
For classifying cognitive activity, the sparse coding algorithm of $L1$
Regularized Learning consistently outperformed 4 variations of ICA across
different numbers of networks and noise levels (p$<$0.001). The NMF algorithms,
which suppressed negative BOLD signal, had the poorest accuracy. Within each
algorithm, encodings using sparser spatial networks (containing more
zero-valued voxels) had higher classification accuracy (p$<$0.001). The success
of sparse coding algorithms may suggest that algorithms which enforce sparse
coding, discourage multitasking, and promote local specialization may capture
better the underlying source processes than those which allow inexhaustible
local processes such as ICA.
|
A Greedy Approach to Adapting the Trace Parameter for Temporal
Difference Learning | One of the main obstacles to broad application of reinforcement learning
methods is the parameter sensitivity of our core learning algorithms. In many
large-scale applications, online computation and function approximation
represent key strategies in scaling up reinforcement learning algorithms. In
this setting, we have effective and reasonably well understood algorithms for
adapting the learning-rate parameter, online during learning. Such
meta-learning approaches can improve robustness of learning and enable
specialization to current task, improving learning speed. For
temporal-difference learning algorithms which we study here, there is yet
another parameter, $\lambda$, that similarly impacts learning speed and
stability in practice. Unfortunately, unlike the learning-rate parameter,
$\lambda$ parametrizes the objective function that temporal-difference methods
optimize. Different choices of $\lambda$ produce different fixed-point
solutions, and thus adapting $\lambda$ online and characterizing the
optimization is substantially more complex than adapting the learning-rate
parameter. There are no meta-learning method for $\lambda$ that can achieve (1)
incremental updating, (2) compatibility with function approximation, and (3)
maintain stability of learning under both on and off-policy sampling. In this
paper we contribute a novel objective function for optimizing $\lambda$ as a
function of state rather than time. We derive a new incremental, linear
complexity $\lambda$-adaption algorithm that does not require offline batch
updating or access to a model of the world, and present a suite of experiments
illustrating the practicality of our new algorithm in three different settings.
Taken together, our contributions represent a concrete step towards black-box
application of temporal-difference learning methods in real world problems.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.