title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Unsupervised Learning using Pretrained CNN and Associative Memory Bank | Deep Convolutional features extracted from a comprehensive labeled dataset,
contain substantial representations which could be effectively used in a new
domain. Despite the fact that generic features achieved good results in many
visual tasks, fine-tuning is required for pretrained deep CNN models to be more
effective and provide state-of-the-art performance. Fine tuning using the
backpropagation algorithm in a supervised setting, is a time and resource
consuming process. In this paper, we present a new architecture and an approach
for unsupervised object recognition that addresses the above mentioned problem
with fine tuning associated with pretrained CNN-based supervised deep learning
approaches while allowing automated feature extraction. Unlike existing works,
our approach is applicable to general object recognition tasks. It uses a
pretrained (on a related domain) CNN model for automated feature extraction
pipelined with a Hopfield network based associative memory bank for storing
patterns for classification purposes. The use of associative memory bank in our
framework allows eliminating backpropagation while providing competitive
performance on an unseen dataset.
|
Covariate Shift Estimation based Adaptive Ensemble Learning for Handling
Non-Stationarity in Motor Imagery related EEG-based Brain-Computer Interface | The non-stationary nature of electroencephalography (EEG) signals makes an
EEG-based brain-computer interface (BCI) a dynamic system, thus improving its
performance is a challenging task. In addition, it is well-known that due to
non-stationarity based covariate shifts, the input data distributions of
EEG-based BCI systems change during inter- and intra-session transitions, which
poses great difficulty for developments of online adaptive data-driven systems.
Ensemble learning approaches have been used previously to tackle this
challenge. However, passive scheme based implementation leads to poor
efficiency while increasing high computational cost. This paper presents a
novel integration of covariate shift estimation and unsupervised adaptive
ensemble learning (CSE-UAEL) to tackle non-stationarity in motor-imagery (MI)
related EEG classification. The proposed method first employs an exponentially
weighted moving average model to detect the covariate shifts in the common
spatial pattern features extracted from MI related brain responses. Then, a
classifier ensemble was created and updated over time to account for changes in
streaming input data distribution wherein new classifiers are added to the
ensemble in accordance with estimated shifts. Furthermore, using two publicly
available BCI-related EEG datasets, the proposed method was extensively
compared with the state-of-the-art single-classifier based passive scheme,
single-classifier based active scheme and ensemble based passive schemes. The
experimental results show that the proposed active scheme based ensemble
learning algorithm significantly enhances the BCI performance in MI
classifications.
|
Alpha-Beta Divergence For Variational Inference | This paper introduces a variational approximation framework using direct
optimization of what is known as the {\it scale invariant Alpha-Beta
divergence} (sAB divergence). This new objective encompasses most variational
objectives that use the Kullback-Leibler, the R{\'e}nyi or the gamma
divergences. It also gives access to objective functions never exploited before
in the context of variational inference. This is achieved via two easy to
interpret control parameters, which allow for a smooth interpolation over the
divergence space while trading-off properties such as mass-covering of a target
distribution and robustness to outliers in the data. Furthermore, the sAB
variational objective can be optimized directly by repurposing existing methods
for Monte Carlo computation of complex variational objectives, leading to
estimates of the divergence instead of variational lower bounds. We show the
advantages of this objective on Bayesian models for regression problems.
|
Large-Scale Unsupervised Deep Representation Learning for Brain
Structure | Machine Learning (ML) is increasingly being used for computer aided diagnosis
of brain related disorders based on structural magnetic resonance imaging (MRI)
data. Most of such work employs biologically and medically meaningful
hand-crafted features calculated from different regions of the brain. The
construction of such highly specialized features requires a considerable amount
of time, manual oversight and careful quality control to ensure the absence of
errors in the computational process. Recent advances in Deep Representation
Learning have shown great promise in extracting highly non-linear and
information-rich features from data. In this paper, we present a novel
large-scale deep unsupervised approach to learn generic feature representations
of structural brain MRI scans, which requires no specialized domain knowledge
or manual intervention. Our method produces low-dimensional representations of
brain structure, which can be used to reconstruct brain images with very low
error and exhibit performance comparable to FreeSurfer features on various
classification tasks.
|
Exploration of Numerical Precision in Deep Neural Networks | Reduced numerical precision is a common technique to reduce computational
cost in many Deep Neural Networks (DNNs). While it has been observed that DNNs
are resilient to small errors and noise, no general result exists that is
capable of predicting a given DNN system architecture's sensitivity to reduced
precision. In this project, we emulate arbitrary bit-width using a specified
floating-point representation with a truncation method, which is applied to the
neural network after each batch. We explore the impact of several model
parameters on the network's training accuracy and show results on the MNIST
dataset. We then present a preliminary theoretical investigation of the error
scaling in both forward and backward propagations. We end with a discussion of
the implications of these results as well as the potential for generalization
to other network architectures.
|
Stack-Pointer Networks for Dependency Parsing | We introduce a novel architecture for dependency parsing: \emph{stack-pointer
networks} (\textbf{\textsc{StackPtr}}). Combining pointer
networks~\citep{vinyals2015pointer} with an internal stack, the proposed model
first reads and encodes the whole sentence, then builds the dependency tree
top-down (from root-to-leaf) in a depth-first fashion. The stack tracks the
status of the depth-first search and the pointer networks select one child for
the word at the top of the stack at each step. The \textsc{StackPtr} parser
benefits from the information of the whole sentence and all previously derived
subtree structures, and removes the left-to-right restriction in classical
transition-based parsers. Yet, the number of steps for building any (including
non-projective) parse tree is linear in the length of the sentence just as
other transition-based parsers, yielding an efficient decoding algorithm with
$O(n^2)$ time complexity. We evaluate our model on 29 treebanks spanning 20
languages and different dependency annotation schemas, and achieve
state-of-the-art performance on 21 of them.
|
A Hierarchical End-to-End Model for Jointly Improving Text Summarization
and Sentiment Classification | Text summarization and sentiment classification both aim to capture the main
ideas of the text but at different levels. Text summarization is to describe
the text within a few sentences, while sentiment classification can be regarded
as a special type of summarization which "summarizes" the text into a even more
abstract fashion, i.e., a sentiment class. Based on this idea, we propose a
hierarchical end-to-end model for joint learning of text summarization and
sentiment classification, where the sentiment classification label is treated
as the further "summarization" of the text summarization output. Hence, the
sentiment classification layer is put upon the text summarization layer, and a
hierarchical structure is derived. Experimental results on Amazon online
reviews datasets show that our model achieves better performance than the
strong baseline systems on both abstractive summarization and sentiment
classification.
|
Local Critic Training of Deep Neural Networks | This paper proposes a novel approach to train deep neural networks by
unlocking the layer-wise dependency of backpropagation training. The approach
employs additional modules called local critic networks besides the main
network model to be trained, which are used to obtain error gradients without
complete feedforward and backward propagation processes. We propose a cascaded
learning strategy for these local networks. In addition, the approach is also
useful from multi-model perspectives, including structural optimization of
neural networks, computationally efficient progressive inference, and ensemble
classification for performance improvement. Experimental results show the
effectiveness of the proposed approach and suggest guidelines for determining
appropriate algorithm parameters.
|
Nonparametric Pricing Analytics with Customer Covariates | Personalized pricing analytics is becoming an essential tool in retailing.
Upon observing the personalized information of each arriving customer, the firm
needs to set a price accordingly based on the covariates such as income,
education background, past purchasing history to extract more revenue. For new
entrants of the business, the lack of historical data may severely limit the
power and profitability of personalized pricing. We propose a nonparametric
pricing policy to simultaneously learn the preference of customers based on the
covariates and maximize the expected revenue over a finite horizon. The policy
does not depend on any prior assumptions on how the personalized information
affects consumers' preferences (such as linear models). It is adaptively splits
the covariate space into smaller bins (hyper-rectangles) and clusters customers
based on their covariates and preferences, offering similar prices for
customers who belong to the same cluster trading off granularity and accuracy.
We show that the algorithm achieves a regret of order $O(\log(T)^2
T^{(2+d)/(4+d)})$, where $T$ is the length of the horizon and $d$ is the
dimension of the covariate. It improves the current regret in the literature
\citep{slivkins2014contextual}, under mild technical conditions in the pricing
context (smoothness and local concavity). We also prove that no policy can
achieve a regret less than $O(T^{(2+d)/(4+d)})$ for a particular instance and
thus demonstrate the near optimality of the proposed policy.
|
Supervector Compression Strategies to Speed up I-Vector System
Development | The front-end factor analysis (FEFA), an extension of principal component
analysis (PPCA) tailored to be used with Gaussian mixture models (GMMs), is
currently the prevalent approach to extract compact utterance-level features
(i-vectors) for automatic speaker verification (ASV) systems. Little research
has been conducted comparing FEFA to the conventional PPCA applied to maximum a
posteriori (MAP) adapted GMM supervectors. We study several alternative
methods, including PPCA, factor analysis (FA), and two supervised approaches,
supervised PPCA (SPPCA) and the recently proposed probabilistic partial least
squares (PPLS), to compress MAP-adapted GMM supervectors. The resulting
i-vectors are used in ASV tasks with a probabilistic linear discriminant
analysis (PLDA) back-end. We experiment on two different datasets, on the
telephone condition of NIST SRE 2010 and on the recent VoxCeleb corpus
collected from YouTube videos containing celebrity interviews recorded in
various acoustical and technical conditions. The results suggest that, in terms
of ASV accuracy, the supervector compression approaches are on a par with FEFA.
The supervised approaches did not result in improved performance. In comparison
to FEFA, we obtained more than hundred-fold (100x) speedups in the total
variability model (TVM) training using the PPCA and FA supervector compression
approaches.
|
Graph Bayesian Optimization: Algorithms, Evaluations and Applications | Network structure optimization is a fundamental task in complex network
analysis. However, almost all the research on Bayesian optimization is aimed at
optimizing the objective functions with vectorial inputs. In this work, we
first present a flexible framework, denoted graph Bayesian optimization, to
handle arbitrary graphs in the Bayesian optimization community. By combining
the proposed framework with graph kernels, it can take full advantage of
implicit graph structural features to supplement explicit features guessed
according to the experience, such as tags of nodes and any attributes of
graphs. The proposed framework can identify which features are more important
during the optimization process. We apply the framework to solve four problems
including two evaluations and two applications to demonstrate its efficacy and
potential applications.
|
SafeRNet: Safe Transportation Routing in the era of Internet of Vehicles
and Mobile Crowd Sensing | World wide road traffic fatality and accident rates are high, and this is
true even in technologically advanced countries like the USA. Despite the
advances in Intelligent Transportation Systems, safe transportation routing
i.e., finding safest routes is largely an overlooked paradigm. In recent years,
large amount of traffic data has been produced by people, Internet of Vehicles
and Internet of Things (IoT). Also, thanks to advances in cloud computing and
proliferation of mobile communication technologies, it is now possible to
perform analysis on vast amount of generated data (crowd sourced) and deliver
the result back to users in real time. This paper proposes SafeRNet, a safe
route computation framework which takes advantage of these technologies to
analyze streaming traffic data and historical data to effectively infer safe
routes and deliver them back to users in real time. SafeRNet utilizes Bayesian
network to formulate safe route model. Furthermore, a case study is presented
to demonstrate the effectiveness of our approach using real traffic data.
SafeRNet intends to improve drivers safety in a modern technology rich
transportation system.
|
Optimization of computational budget for power system risk assessment | We address the problem of maintaining high voltage power transmission
networks in security at all time, namely anticipating exceeding of thermal
limit for eventual single line disconnection (whatever its cause may be) by
running slow, but accurate, physical grid simulators. New conceptual frameworks
are calling for a probabilistic risk-based security criterion. However, these
approaches suffer from high requirements in terms of tractability. Here, we
propose a new method to assess the risk. This method uses both machine learning
techniques (artificial neural networks) and more standard simulators based on
physical laws. More specifically we train neural networks to estimate the
overall dangerousness of a grid state. A classical benchmark problem (manpower
118 buses test case) is used to show the strengths of the proposed method.
|
Disentangling Language and Knowledge in Task-Oriented Dialogs | The Knowledge Base (KB) used for real-world applications, such as booking a
movie or restaurant reservation, keeps changing over time. End-to-end neural
networks trained for these task-oriented dialogs are expected to be immune to
any changes in the KB. However, existing approaches breakdown when asked to
handle such changes. We propose an encoder-decoder architecture (BoSsNet) with
a novel Bag-of-Sequences (BoSs) memory, which facilitates the disentangled
learning of the response's language model and its knowledge incorporation.
Consequently, the KB can be modified with new knowledge without a drop in
interpretability. We find that BoSsNet outperforms state-of-the-art models,
with considerable improvements (> 10\%) on bAbI OOV test sets and other
human-human datasets. We also systematically modify existing datasets to
measure disentanglement and show BoSsNet to be robust to KB modifications.
|
audEERING's approach to the One-Minute-Gradual Emotion Challenge | This paper describes audEERING's submissions as well as additional
evaluations for the One-Minute-Gradual (OMG) emotion recognition challenge. We
provide the results for audio and video processing on subject (in)dependent
evaluations. On the provided Development set, we achieved 0.343 Concordance
Correlation Coefficient (CCC) for arousal (from audio) and .401 for valence
(from video).
|
An Asymptotically Optimal Strategy for Constrained Multi-armed Bandit
Problems | For the stochastic multi-armed bandit (MAB) problem from a constrained model
that generalizes the classical one, we show that an asymptotic optimality is
achievable by a simple strategy extended from the $\epsilon_t$-greedy strategy.
We provide a finite-time lower bound on the probability of correct selection of
an optimal near-feasible arm that holds for all time steps. Under some
conditions, the bound approaches one as time $t$ goes to infinity. A particular
example sequence of $\{\epsilon_t\}$ having the asymptotic convergence rate in
the order of $(1-\frac{1}{t})^4$ that holds from a sufficiently large $t$ is
also discussed.
|
Improving a Neural Semantic Parser by Counterfactual Learning from Human
Bandit Feedback | Counterfactual learning from human bandit feedback describes a scenario where
user feedback on the quality of outputs of a historic system is logged and used
to improve a target system. We show how to apply this learning framework to
neural semantic parsing. From a machine learning perspective, the key challenge
lies in a proper reweighting of the estimator so as to avoid known degeneracies
in counterfactual learning, while still being applicable to stochastic gradient
optimization. To conduct experiments with human users, we devise an easy-to-use
interface to collect human feedback on semantic parses. Our work is the first
to show that semantic parsers can be improved significantly by counterfactual
learning from logged human feedback data.
|
Learning Pretopological Spaces to Model Complex Propagation Phenomena: A
Multiple Instance Learning Approach Based on a Logical Modeling | This paper addresses the problem of learning the concept of "propagation" in
the pretopology theoretical formalism. Our proposal is first to define the
pseudo-closure operator (modeling the propagation concept) as a logical
combination of neighborhoods. We show that learning such an operator lapses
into the Multiple Instance (MI) framework, where the learning process is
performed on bags of instances instead of individual instances. Though this
framework is well suited for this task, its use for learning a pretopological
space leads to a set of bags exponential in size. To overcome this issue we
thus propose a learning method based on a low estimation of the bags covered by
a concept under construction. As an experiment, percolation processes (forest
fires typically) are simulated and the corresponding propagation models are
learned based on a subset of observations. It reveals that the proposed MI
approach is significantly more efficient on the task of propagation model
recognition than existing methods.
|
Semantic Channel and Shannon's Channel Mutually Match for Multi-Label
Classification | A group of transition probability functions form a Shannon's channel whereas
a group of truth functions form a semantic channel. Label learning is to let
semantic channels match Shannon's channels and label selection is to let
Shannon's channels match semantic channels. The Channel Matching (CM) algorithm
is provided for multi-label classification. This algorithm adheres to maximum
semantic information criterion which is compatible with maximum likelihood
criterion and regularized least squares criterion. If samples are very large,
we can directly convert Shannon's channels into semantic channels by the third
kind of Bayes' theorem; otherwise, we can train truth functions with parameters
by sampling distributions. A label may be a Boolean function of some atomic
labels. For simplifying learning, we may only obtain the truth functions of
some atomic label. For a given label, instances are divided into three kinds
(positive, negative, and unclear) instead of two kinds as in popular studies so
that the problem with binary relevance is avoided. For each instance, the
classifier selects a compound label with most semantic information or richest
connotation. As a predictive model, the semantic channel does not change with
the prior probability distribution (source) of instances. It still works when
the source is changed. The classifier changes with the source, and hence can
overcome class-imbalance problem. It is shown that the old population's
increasing will change the classifier for label "Old" and has been impelling
the semantic evolution of "Old". The CM iteration algorithm for unseen instance
classification is introduced.
|
Spiking Deep Residual Network | Spiking neural networks (SNNs) have received significant attention for their
biological plausibility. SNNs theoretically have at least the same
computational power as traditional artificial neural networks (ANNs). They
possess potential of achieving energy-efficiency while keeping comparable
performance to deep neural networks (DNNs). However, it is still a big
challenge to train a very deep SNN. In this paper, we propose an efficient
approach to build a spiking version of deep residual network (ResNet). ResNet
is considered as a kind of the state-of-the-art convolutional neural networks
(CNNs). We employ the idea of converting a trained ResNet to a network of
spiking neurons, named Spiking ResNet (S-ResNet). We propose a shortcut
conversion model to appropriately scale continuous-valued activations to match
firing rates in SNN, and a compensation mechanism to reduce the error caused by
discretisation. Experimental results demonstrate that, compared with the
state-of-the-art SNN approaches, the proposed Spiking ResNet achieves the best
performance on CIFAR-10, CIFAR-100, and ImageNet 2012. Our work is the first
time to build a SNN deeper than 40, with comparable performance to ANNs on a
large-scale dataset.
|
Boosting Noise Robustness of Acoustic Model via Deep Adversarial
Training | In realistic environments, speech is usually interfered by various noise and
reverberation, which dramatically degrades the performance of automatic speech
recognition (ASR) systems. To alleviate this issue, the commonest way is to use
a well-designed speech enhancement approach as the front-end of ASR. However,
more complex pipelines, more computations and even higher hardware costs
(microphone array) are additionally consumed for this kind of methods. In
addition, speech enhancement would result in speech distortions and mismatches
to training. In this paper, we propose an adversarial training method to
directly boost noise robustness of acoustic model. Specifically, a jointly
compositional scheme of generative adversarial net (GAN) and neural
network-based acoustic model (AM) is used in the training phase. GAN is used to
generate clean feature representations from noisy features by the guidance of a
discriminator that tries to distinguish between the true clean signals and
generated signals. The joint optimization of generator, discriminator and AM
concentrates the strengths of both GAN and AM for speech recognition.
Systematic experiments on CHiME-4 show that the proposed method significantly
improves the noise robustness of AM and achieves the average relative error
rate reduction of 23.38% and 11.54% on the development and test set,
respectively.
|
Anomaly and Change Detection in Graph Streams through Constant-Curvature
Manifold Embeddings | Mapping complex input data into suitable lower dimensional manifolds is a
common procedure in machine learning. This step is beneficial mainly for two
reasons: (1) it reduces the data dimensionality and (2) it provides a new data
representation possibly characterised by convenient geometric properties.
Euclidean spaces are by far the most widely used embedding spaces, thanks to
their well-understood structure and large availability of consolidated
inference methods. However, recent research demonstrated that many types of
complex data (e.g., those represented as graphs) are actually better described
by non-Euclidean geometries. Here, we investigate how embedding graphs on
constant-curvature manifolds (hyper-spherical and hyperbolic manifolds) impacts
on the ability to detect changes in sequences of attributed graphs. The
proposed methodology consists in embedding graphs into a geometric space and
perform change detection there by means of conventional methods for numerical
streams. The curvature of the space is a parameter that we learn to reproduce
the geometry of the original application-dependent graph space. Preliminary
experimental results show the potential capability of representing graphs by
means of curved manifold, in particular for change and anomaly detection
problems.
|
Open Loop Execution of Tree-Search Algorithms, extended version | In the context of tree-search stochastic planning algorithms where a
generative model is available, we consider on-line planning algorithms building
trees in order to recommend an action. We investigate the question of avoiding
re-planning in subsequent decision steps by directly using sub-trees as action
recommender. Firstly, we propose a method for open loop control via a new
algorithm taking the decision of re-planning or not at each time step based on
an analysis of the statistics of the sub-tree. Secondly, we show that the
probability of selecting a suboptimal action at any depth of the tree can be
upper bounded and converges towards zero. Moreover, this upper bound decays in
a logarithmic way between subsequent depths. This leads to a distinction
between node-wise optimality and state-wise optimality. Finally, we empirically
demonstrate that our method achieves a compromise between loss of performance
and computational gain.
|
Siamese networks for generating adversarial examples | Machine learning models are vulnerable to adversarial examples. An adversary
modifies the input data such that humans still assign the same label, however,
machine learning models misclassify it. Previous approaches in the literature
demonstrated that adversarial examples can even be generated for the remotely
hosted model. In this paper, we propose a Siamese network based approach to
generate adversarial examples for a multiclass target CNN. We assume that the
adversary do not possess any knowledge of the target data distribution, and we
use an unlabeled mismatched dataset to query the target, e.g., for the
ResNet-50 target, we use the Food-101 dataset as the query. Initially, the
target model assigns labels to the query dataset, and a Siamese network is
trained on the image pairs derived from these multiclass labels. We learn the
\emph{adversarial perturbations} for the Siamese model and show that these
perturbations are also adversarial w.r.t. the target model. In experimental
results, we demonstrate effectiveness of our approach on MNIST, CIFAR-10 and
ImageNet targets with TinyImageNet/Food-101 query datasets.
|
Noisin: Unbiased Regularization for Recurrent Neural Networks | Recurrent neural networks (RNNs) are powerful models of sequential data. They
have been successfully used in domains such as text and speech. However, RNNs
are susceptible to overfitting; regularization is important. In this paper we
develop Noisin, a new method for regularizing RNNs. Noisin injects random noise
into the hidden states of the RNN and then maximizes the corresponding marginal
likelihood of the data. We show how Noisin applies to any RNN and we study many
different types of noise. Noisin is unbiased--it preserves the underlying RNN
on average. We characterize how Noisin regularizes its RNN both theoretically
and empirically. On language modeling benchmarks, Noisin improves over dropout
by as much as 12.2% on the Penn Treebank and 9.4% on the Wikitext-2 dataset. We
also compared the state-of-the-art language model of Yang et al. 2017, both
with and without Noisin. On the Penn Treebank, the method with Noisin more
quickly reaches state-of-the-art performance.
|
RECS: Robust Graph Embedding Using Connection Subgraphs | The success of graph embeddings or node representation learning in a variety
of downstream tasks, such as node classification, link prediction, and
recommendation systems, has led to their popularity in recent years.
Representation learning algorithms aim to preserve local and global network
structure by identifying node neighborhood notions. However, many existing
algorithms generate embeddings that fail to properly preserve the network
structure, or lead to unstable representations due to random processes (e.g.,
random walks to generate context) and, thus, cannot generate to multi-graph
problems. In this paper, we propose RECS, a novel, stable graph embedding
algorithmic framework. RECS learns graph representations using connection
subgraphs by employing the analogy of graphs with electrical circuits. It
preserves both local and global connectivity patterns, and addresses the issue
of high-degree nodes. Further, it exploits the strength of weak ties and
meta-data that have been neglected by baselines. The experiments show that RECS
outperforms state-of-the-art algorithms by up to 36.85% on multi-label
classification problem. Further, in contrast to baselines, RECS, being
deterministic, is completely stable.
|
How deep should be the depth of convolutional neural networks: a
backyard dog case study | The work concerns the problem of reducing a pre-trained deep neuronal network
to a smaller network, with just few layers, whilst retaining the network's
functionality on a given task
The proposed approach is motivated by the observation that the aim to deliver
the highest accuracy possible in the broadest range of operational conditions,
which many deep neural networks models strive to achieve, may not necessarily
be always needed, desired, or even achievable due to the lack of data or
technical constraints. In relation to the face recognition problem, we
formulated an example of such a usecase, the `backyard dog' problem. The
`backyard dog', implemented by a lean network, should correctly identify
members from a limited group of individuals, a `family', and should distinguish
between them. At the same time, the network must produce an alarm to an image
of an individual who is not in a member of the family. To produce such a
network, we propose a shallowing algorithm. The algorithm takes an existing
deep learning model on its input and outputs a shallowed version of it. The
algorithm is non-iterative and is based on the Advanced Supervised Principal
Component Analysis. Performance of the algorithm is assessed in exhaustive
numerical experiments. In the above usecase, the `backyard dog' problem, the
method is capable of drastically reducing the depth of deep learning neural
networks, albeit at the cost of mild performance deterioration.
We developed a simple non-iterative method for shallowing down pre-trained
deep networks. The method is generic in the sense that it applies to a broad
class of feed-forward networks, and is based on the Advanced Supervise
Principal Component Analysis. The method enables generation of families of
smaller-size shallower specialized networks tuned for specific operational
conditions and tasks from a single larger and more universal legacy network.
|
Lifted Neural Networks | We describe a novel family of models of multi- layer feedforward neural
networks in which the activation functions are encoded via penalties in the
training problem. Our approach is based on representing a non-decreasing
activation function as the argmin of an appropriate convex optimiza- tion
problem. The new framework allows for algo- rithms such as block-coordinate
descent methods to be applied, in which each step is composed of a simple (no
hidden layer) supervised learning problem that is parallelizable across data
points and/or layers. Experiments indicate that the pro- posed models provide
excellent initial guesses for weights for standard neural networks. In addi-
tion, the model provides avenues for interesting extensions, such as robustness
against noisy in- puts and optimizing over parameters in activation functions.
|
A Deep Learning Model with Hierarchical LSTMs and Supervised Attention
for Anti-Phishing | Anti-phishing aims to detect phishing content/documents in a pool of textual
data. This is an important problem in cybersecurity that can help to guard
users from fraudulent information. Natural language processing (NLP) offers a
natural solution for this problem as it is capable of analyzing the textual
content to perform intelligent recognition. In this work, we investigate
state-of-the-art techniques for text categorization in NLP to address the
problem of anti-phishing for emails (i.e, predicting if an email is phishing or
not). These techniques are based on deep learning models that have attracted
much attention from the community recently. In particular, we present a
framework with hierarchical long short-term memory networks (H-LSTMs) and
attention mechanisms to model the emails simultaneously at the word and the
sentence level. Our expectation is to produce an effective model for
anti-phishing and demonstrate the effectiveness of deep learning for problems
in cybersecurity.
|
Distribution Assertive Regression | In regression modelling approach, the main step is to fit the regression line
as close as possible to the target variable. In this process most algorithms
try to fit all of the data in a single line and hence fitting all parts of
target variable in one go. It was observed that the error between predicted and
target variable usually have a varying behavior across the various quantiles of
the dependent variable and hence single point diagnostic like MAPE has its
limitation to signify the level of fitness across the distribution of
Y(dependent variable). To address this problem, a novel approach is proposed in
the paper to deal with regression fitting over various quantiles of target
variable. Using this approach we have significantly improved the eccentric
behavior of the distance (error) between predicted and actual value of
regression. Our proposed solution is based on understanding the segmented
behavior of the data with respect to the internal segments within the data and
approach for retrospectively fitting the data based on each quantile behavior.
We believe exploring and using this approach would help in achieving better and
more explainable results in most settings of real world data modelling
problems.
|
Estimating Learnability in the Sublinear Data Regime | We consider the problem of estimating how well a model class is capable of
fitting a distribution of labeled data. We show that it is often possible to
accurately estimate this "learnability" even when given an amount of data that
is too small to reliably learn any accurate model. Our first result applies to
the setting where the data is drawn from a $d$-dimensional distribution with
isotropic covariance (or known covariance), and the label of each datapoint is
an arbitrary noisy function of the datapoint. In this setting, we show that
with $O(\sqrt{d})$ samples, one can accurately estimate the fraction of the
variance of the label that can be explained via the best linear function of the
data. In contrast to this sublinear sample size, finding an approximation of
the best-fit linear function requires on the order of $d$ samples. Our
sublinear sample results and approach also extend to the non-isotropic setting,
where the data distribution has an (unknown) arbitrary covariance matrix: we
show that, if the label $y$ of point $x$ is a linear function with independent
noise, $y = \langle x , \beta \rangle + noise$ with $\|\beta \|$ bounded, the
variance of the noise can be estimated to error $\epsilon$ with
$O(d^{1-1/\log{1/\epsilon}})$ if the covariance matrix has bounded condition
number, or $O(d^{1-\sqrt{\epsilon}})$ if there are no bounds on the condition
number. We also establish that these sample complexities are optimal, to
constant factors. Finally, we extend these techniques to the setting of binary
classification, where we obtain analogous sample complexities for the problem
of estimating the prediction error of the best linear classifier, in a natural
model of binary labeled data. We demonstrate the practical viability of our
approaches on several real and synthetic datasets.
|
BelMan: Bayesian Bandits on the Belief--Reward Manifold | We propose a generic, Bayesian, information geometric approach to the
exploration--exploitation trade-off in multi-armed bandit problems. Our
approach, BelMan, uniformly supports pure exploration,
exploration--exploitation, and two-phase bandit problems. The knowledge on
bandit arms and their reward distributions is summarised by the barycentre of
the joint distributions of beliefs and rewards of the arms, the
\emph{pseudobelief-reward}, within the beliefs-rewards manifold. BelMan
alternates \emph{information projection} and \emph{reverse information
projection}, i.e., projection of the pseudobelief-reward onto beliefs-rewards
to choose the arm to play, and projection of the resulting beliefs-rewards onto
the pseudobelief-reward. It introduces a mechanism that infuses an exploitative
bias by means of a \emph{focal distribution}, i.e., a reward distribution that
gradually concentrates on higher rewards. Comparative performance evaluation
with state-of-the-art algorithms shows that BelMan is not only competitive but
can also outperform other approaches in specific setups, for instance involving
many arms and continuous rewards.
|
Sharp convergence rates for Langevin dynamics in the nonconvex setting | We study the problem of sampling from a distribution $p^*(x) \propto
\exp\left(-U(x)\right)$, where the function $U$ is $L$-smooth everywhere and
$m$-strongly convex outside a ball of radius $R$, but potentially nonconvex
inside this ball. We study both overdamped and underdamped Langevin MCMC and
establish upper bounds on the number of steps required to obtain a sample from
a distribution that is within $\epsilon$ of $p^*$ in $1$-Wasserstein distance.
For the first-order method (overdamped Langevin MCMC), the iteration complexity
is $\tilde{\mathcal{O}}\left(e^{cLR^2}d/\epsilon^2\right)$, where $d$ is the
dimension of the underlying space. For the second-order method (underdamped
Langevin MCMC), the iteration complexity is
$\tilde{\mathcal{O}}\left(e^{cLR^2}\sqrt{d}/\epsilon\right)$ for an explicit
positive constant $c$. Surprisingly, the iteration complexity for both these
algorithms is only polynomial in the dimension $d$ and the target accuracy
$\epsilon$. It is exponential, however, in the problem parameter $LR^2$, which
is a measure of non-log-concavity of the target distribution.
|
Intracranial Error Detection via Deep Learning | Deep learning techniques have revolutionized the field of machine learning
and were recently successfully applied to various classification problems in
noninvasive electroencephalography (EEG). However, these methods were so far
only rarely evaluated for use in intracranial EEG. We employed convolutional
neural networks (CNNs) to classify and characterize the error-related brain
response as measured in 24 intracranial EEG recordings. Decoding accuracies of
CNNs were significantly higher than those of a regularized linear discriminant
analysis. Using time-resolved deep decoding, it was possible to classify errors
in various regions in the human brain, and further to decode errors over 200 ms
before the actual erroneous button press, e.g., in the precentral gyrus.
Moreover, deeper networks performed better than shallower networks in
distinguishing correct from error trials in all-channel decoding. In single
recordings, up to 100 % decoding accuracy was achieved. Visualization of the
networks' learned features indicated that multivariate decoding on an ensemble
of channels yields related, albeit non-redundant information compared to
single-channel decoding. In summary, here we show the usefulness of deep
learning for both intracranial error decoding and mapping of the
spatio-temporal structure of the human error processing network.
|
Combinatorial Pure Exploration with Continuous and Separable Reward
Functions and Its Applications (Extended Version) | We study the Combinatorial Pure Exploration problem with Continuous and
Separable reward functions (CPE-CS) in the stochastic multi-armed bandit
setting. In a CPE-CS instance, we are given several stochastic arms with
unknown distributions, as well as a collection of possible decisions. Each
decision has a reward according to the distributions of arms. The goal is to
identify the decision with the maximum reward, using as few arm samples as
possible. The problem generalizes the combinatorial pure exploration problem
with linear rewards, which has attracted significant attention in recent years.
In this paper, we propose an adaptive learning algorithm for the CPE-CS
problem, and analyze its sample complexity. In particular, we introduce a new
hardness measure called the consistent optimality hardness, and give both the
upper and lower bounds of sample complexity. Moreover, we give examples to
demonstrate that our solution has the capacity to deal with non-linear reward
functions.
|
Beyond the Click-Through Rate: Web Link Selection with Multi-level
Feedback | The web link selection problem is to select a small subset of web links from
a large web link pool, and to place the selected links on a web page that can
only accommodate a limited number of links, e.g., advertisements,
recommendations, or news feeds. Despite the long concerned click-through rate
which reflects the attractiveness of the link itself, the revenue can only be
obtained from user actions after clicks, e.g., purchasing after being directed
to the product pages by recommendation links. Thus, the web links have an
intrinsic \emph{multi-level feedback structure}. With this observation, we
consider the context-free web link selection problem, where the objective is to
maximize revenue while ensuring that the attractiveness is no less than a
preset threshold. The key challenge of the problem is that each link's
multi-level feedbacks are stochastic, and unobservable unless the link is
selected. We model this problem with a constrained stochastic multi-armed
bandit formulation, and design an efficient link selection algorithm, called
Constrained Upper Confidence Bound algorithm (\textbf{Con-UCB}), and prove
$O(\sqrt{T\ln T})$ bounds on both the regret and the violation of the
attractiveness constraint. We conduct extensive experiments on three real-world
datasets, and show that \textbf{Con-UCB} outperforms state-of-the-art
context-free bandit algorithms concerning the multi-level feedback structure.
|
A brief introduction to the Grey Machine Learning | This paper presents a brief introduction to the key points of the Grey
Machine Learning (GML) based on the kernels. The general formulation of the
grey system models have been firstly summarized, and then the nonlinear
extension of the grey models have been developed also with general
formulations. The kernel implicit mapping is used to estimate the nonlinear
function of the GML model, by extending the nonparametric formulation of the
LSSVM, the estimation of the nonlinear function of the GML model can also be
expressed by the kernels. A short discussion on the priority of this new
framework to the existing grey models and LSSVM have also been discussed in
this paper. And the perspectives and future orientations of this framework have
also been presented.
|
Dynamic Control Flow in Large-Scale Machine Learning | Many recent machine learning models rely on fine-grained dynamic control flow
for training and inference. In particular, models based on recurrent neural
networks and on reinforcement learning depend on recurrence relations,
data-dependent conditional execution, and other features that call for dynamic
control flow. These applications benefit from the ability to make rapid
control-flow decisions across a set of computing devices in a distributed
system. For performance, scalability, and expressiveness, a machine learning
system must support dynamic control flow in distributed and heterogeneous
environments.
This paper presents a programming model for distributed machine learning that
supports dynamic control flow. We describe the design of the programming model,
and its implementation in TensorFlow, a distributed machine learning system.
Our approach extends the use of dataflow graphs to represent machine learning
models, offering several distinctive features. First, the branches of
conditionals and bodies of loops can be partitioned across many machines to run
on a set of heterogeneous devices, including CPUs, GPUs, and custom ASICs.
Second, programs written in our model support automatic differentiation and
distributed gradient computations, which are necessary for training machine
learning models that use control flow. Third, our choice of non-strict
semantics enables multiple loop iterations to execute in parallel across
machines, and to overlap compute and I/O operations.
We have done our work in the context of TensorFlow, and it has been used
extensively in research and production. We evaluate it using several real-world
applications, and demonstrate its performance and scalability.
|
Towards a Spectrum of Graph Convolutional Networks | We present our ongoing work on understanding the limitations of graph
convolutional networks (GCNs) as well as our work on generalizations of graph
convolutions for representing more complex node attribute dependencies. Based
on an analysis of GCNs with the help of the corresponding computation graphs,
we propose a generalization of existing GCNs where the aggregation operations
are (a) determined by structural properties of the local neighborhood graphs
and (b) not restricted to weighted averages. We show that the proposed approach
is strictly more expressive while requiring only a modest increase in the
number of parameters and computations. We also show that the proposed
generalization is identical to standard convolutional layers when applied to
regular grid graphs.
|
Inference for $L_2$-Boosting | We propose a statistical inference framework for the component-wise
functional gradient descent algorithm (CFGD) under normality assumption for
model errors, also known as $L_2$-Boosting. The CFGD is one of the most
versatile tools to analyze data, because it scales well to high-dimensional
data sets, allows for a very flexible definition of additive regression models
and incorporates inbuilt variable selection. Due to the variable selection, we
build on recent proposals for post-selection inference. However, the iterative
nature of component-wise boosting, which can repeatedly select the same
component to update, necessitates adaptations and extensions to existing
approaches. We propose tests and confidence intervals for linear, grouped and
penalized additive model components selected by $L_2$-Boosting. Our concepts
also transfer to slow-learning algorithms more generally, and to other
selection techniques which restrict the response space to more complex sets
than polyhedra. We apply our framework to an additive model for sales prices of
residential apartments and investigate the properties of our concepts in
simulation studies.
|
Bayesian active learning for choice models with deep Gaussian processes | In this paper, we propose an active learning algorithm and models which can
gradually learn individual's preference through pairwise comparisons. The
active learning scheme aims at finding individual's most preferred choice with
minimized number of pairwise comparisons. The pairwise comparisons are encoded
into probabilistic models based on assumptions of choice models and deep
Gaussian processes. The next-to-compare decision is determined by a novel
acquisition function. We benchmark the proposed algorithm and models using
functions with multiple local optima and one public airline itinerary dataset.
The experiments indicate the effectiveness of our active learning algorithm and
models.
|
t-PINE: Tensor-based Predictable and Interpretable Node Embeddings | Graph representations have increasingly grown in popularity during the last
years. Existing representation learning approaches explicitly encode network
structure. Despite their good performance in downstream processes (e.g., node
classification, link prediction), there is still room for improvement in
different aspects, like efficacy, visualization, and interpretability. In this
paper, we propose, t-PINE, a method that addresses these limitations. Contrary
to baseline methods, which generally learn explicit graph representations by
solely using an adjacency matrix, t-PINE avails a multi-view information graph,
the adjacency matrix represents the first view, and a nearest neighbor
adjacency, computed over the node features, is the second view, in order to
learn explicit and implicit node representations, using the Canonical Polyadic
(a.k.a. CP) decomposition. We argue that the implicit and the explicit mapping
from a higher-dimensional to a lower-dimensional vector space is the key to
learn more useful, highly predictable, and gracefully interpretable
representations. Having good interpretable representations provides a good
guidance to understand how each view contributes to the representation learning
process. In addition, it helps us to exclude unrelated dimensions. Extensive
experiments show that t-PINE drastically outperforms baseline methods by up to
158.6% with respect to Micro-F1, in several multi-label classification
problems, while it has high visualization and interpretability utility.
|
RMDL: Random Multimodel Deep Learning for Classification | The continually increasing number of complex datasets each year necessitates
ever improving machine learning methods for robust and accurate categorization
of these data. This paper introduces Random Multimodel Deep Learning (RMDL): a
new ensemble, deep learning approach for classification. Deep learning models
have achieved state-of-the-art results across many domains. RMDL solves the
problem of finding the best deep learning structure and architecture while
simultaneously improving robustness and accuracy through ensembles of deep
learning architectures. RDML can accept as input a variety data to include
text, video, images, and symbolic. This paper describes RMDL and shows test
results for image and text data including MNIST, CIFAR-10, WOS, Reuters, IMDB,
and 20newsgroup. These test results show that RDML produces consistently better
performance than standard methods over a broad range of data types and
classification problems.
|
Power Law in Sparsified Deep Neural Networks | The power law has been observed in the degree distributions of many
biological neural networks. Sparse deep neural networks, which learn an
economical representation from the data, resemble biological neural networks in
many ways. In this paper, we study if these artificial networks also exhibit
properties of the power law. Experimental results on two popular deep learning
models, namely, multilayer perceptrons and convolutional neural networks, are
affirmative. The power law is also naturally related to preferential
attachment. To study the dynamical properties of deep networks in continual
learning, we propose an internal preferential attachment model to explain how
the network topology evolves. Experimental results show that with the arrival
of a new task, the new connections made follow this preferential attachment
process.
|
Exploration by Distributional Reinforcement Learning | We propose a framework based on distributional reinforcement learning and
recent attempts to combine Bayesian parameter updates with deep reinforcement
learning. We show that our proposed framework conceptually unifies multiple
previous methods in exploration. We also derive a practical algorithm that
achieves efficient exploration on challenging control tasks.
|
Enhancing the Regularization Effect of Weight Pruning in Artificial
Neural Networks | Artificial neural networks (ANNs) may not be worth their computational/memory
costs when used in mobile phones or embedded devices. Parameter-pruning
algorithms combat these costs, with some algorithms capable of removing over
90% of an ANN's weights without harming the ANN's performance. Removing weights
from an ANN is a form of regularization, but existing pruning algorithms do not
significantly improve generalization error. We show that pruning ANNs can
improve generalization if pruning targets large weights instead of small
weights. Applying our pruning algorithm to an ANN leads to a higher image
classification accuracy on CIFAR-10 data than applying the popular regularizer
dropout. The pruning couples this higher accuracy with an 85% reduction of the
ANN's parameter count.
|
Learning to See in the Dark | Imaging in low light is challenging due to low photon count and low SNR.
Short-exposure images suffer from noise, while long exposure can induce blur
and is often impractical. A variety of denoising, deblurring, and enhancement
techniques have been proposed, but their effectiveness is limited in extreme
conditions, such as video-rate imaging at night. To support the development of
learning-based pipelines for low-light image processing, we introduce a dataset
of raw short-exposure low-light images, with corresponding long-exposure
reference images. Using the presented dataset, we develop a pipeline for
processing low-light images, based on end-to-end training of a
fully-convolutional network. The network operates directly on raw sensor data
and replaces much of the traditional image processing pipeline, which tends to
perform poorly on such data. We report promising results on the new dataset,
analyze factors that affect performance, and highlight opportunities for future
work. The results are shown in the supplementary video at
https://youtu.be/qWKUFK7MWvg
|
Improve Uncertainty Estimation for Unknown Classes in Bayesian Neural
Networks with Semi-Supervised /One Set Classification | Although deep neural network (DNN) has achieved many state-of-the-art
results, estimating the uncertainty presented in the DNN model and the data is
a challenging task. Problems related to uncertainty such as classifying unknown
classes (class which does not appear in the training data) data as known class
with high confidence, is critically concerned in the safety domain area (e.g,
autonomous driving, medical diagnosis). In this paper, we show that applying
current Bayesian Neural Network (BNN) techniques alone does not effectively
capture the uncertainty. To tackle this problem, we introduce a simple way to
improve the BNN by using one class classification (in this paper, we use the
term "set classification" instead). We empirically show the result of our
method on an experiment which involves three datasets: MNIST, notMNIST and
FMNIST.
|
Motion Planning Among Dynamic, Decision-Making Agents with Deep
Reinforcement Learning | Robots that navigate among pedestrians use collision avoidance algorithms to
enable safe and efficient operation. Recent works present deep reinforcement
learning as a framework to model the complex interactions and cooperation.
However, they are implemented using key assumptions about other agents'
behavior that deviate from reality as the number of agents in the environment
increases. This work extends our previous approach to develop an algorithm that
learns collision avoidance among a variety of types of dynamic agents without
assuming they follow any particular behavior rules. This work also introduces a
strategy using LSTM that enables the algorithm to use observations of an
arbitrary number of other agents, instead of previous methods that have a fixed
observation size. The proposed algorithm outperforms our previous approach in
simulation as the number of agents increases, and the algorithm is demonstrated
on a fully autonomous robotic vehicle traveling at human walking speed, without
the use of a 3D Lidar.
|
Unsupervised Feature Learning via Non-Parametric Instance-level
Discrimination | Neural net classifiers trained on data with annotated class labels can also
capture apparent visual similarity among categories without being directed to
do so. We study whether this observation can be extended beyond the
conventional domain of supervised learning: Can we learn a good feature
representation that captures apparent similarity among instances, instead of
classes, by merely asking the feature to be discriminative of individual
instances? We formulate this intuition as a non-parametric classification
problem at the instance-level, and use noise-contrastive estimation to tackle
the computational challenges imposed by the large number of instance classes.
Our experimental results demonstrate that, under unsupervised learning
settings, our method surpasses the state-of-the-art on ImageNet classification
by a large margin. Our method is also remarkable for consistently improving
test performance with more training data and better network architectures. By
fine-tuning the learned feature, we further obtain competitive results for
semi-supervised learning and object detection tasks. Our non-parametric model
is highly compact: With 128 features per image, our method requires only 600MB
storage for a million images, enabling fast nearest neighbour retrieval at the
run time.
|
Transfer Learning of Artist Group Factors to Musical Genre
Classification | The automated recognition of music genres from audio information is a
challenging problem, as genre labels are subjective and noisy. Artist labels
are less subjective and less noisy, while certain artists may relate more
strongly to certain genres. At the same time, at prediction time, it is not
guaranteed that artist labels are available for a given audio segment.
Therefore, in this work, we propose to apply the transfer learning framework,
learning artist-related information which will be used at inference time for
genre classification. We consider different types of artist-related
information, expressed through artist group factors, which will allow for more
efficient learning and stronger robustness to potential label noise.
Furthermore, we investigate how to achieve the highest validation accuracy on
the given FMA dataset, by experimenting with various kinds of transfer methods,
including single-task transfer, multi-task transfer and finally multi-task
learning.
|
Deep Reinforcement Learning for Playing 2.5D Fighting Games | Deep reinforcement learning has shown its success in game playing. However,
2.5D fighting games would be a challenging task to handle due to ambiguity in
visual appearances like height or depth of the characters. Moreover, actions in
such games typically involve particular sequential action orders, which also
makes the network design very difficult. Based on the network of Asynchronous
Advantage Actor-Critic (A3C), we create an OpenAI-gym-like gaming environment
with the game of Little Fighter 2 (LF2), and present a novel A3C+ network for
learning RL agents. The introduced model includes a Recurrent Info network,
which utilizes game-related info features with recurrent layers to observe
combo skills for fighting. In the experiments, we consider LF2 in different
settings, which successfully demonstrates the use of our proposed model for
learning 2.5D fighting games.
|
A Constraint-Based Algorithm For Causal Discovery with Cycles, Latent
Variables and Selection Bias | Causal processes in nature may contain cycles, and real datasets may violate
causal sufficiency as well as contain selection bias. No constraint-based
causal discovery algorithm can currently handle cycles, latent variables and
selection bias (CLS) simultaneously. I therefore introduce an algorithm called
Cyclic Causal Inference (CCI) that makes sound inferences with a conditional
independence oracle under CLS, provided that we can represent the cyclic causal
process as a non-recursive linear structural equation model with independent
errors. Empirical results show that CCI outperforms CCD in the cyclic case as
well as rivals FCI and RFCI in the acyclic case.
|
Developing parsimonious ensembles using ensemble diversity within a
reinforcement learning framework | Heterogeneous ensembles built from the predictions of a wide variety and
large number of diverse base predictors represent a potent approach to building
predictive models for problems where the ideal base/individual predictor may
not be obvious. Ensemble selection is an especially promising approach here,
not only for improving prediction performance, but also because of its ability
to select a collectively predictive subset, often a relatively small one, of
the base predictors. In this paper, we present a set of algorithms that
explicitly incorporate ensemble diversity, a known factor influencing
predictive performance of ensembles, into a reinforcement learning framework
for ensemble selection. We rigorously tested these approaches on several
challenging problems and associated data sets, yielding that several of them
produced more accurate ensembles than those that don't explicitly consider
diversity. More importantly, these diversity-incorporating ensembles were much
smaller in size, i.e., more parsimonious, than the latter types of ensembles.
This can eventually aid the interpretation or reverse engineering of predictive
models assimilated into the resultant ensemble(s).
|
Population Anomaly Detection through Deep Gaussianization | We introduce an algorithmic method for population anomaly detection based on
gaussianization through an adversarial autoencoder. This method is applicable
to detection of `soft' anomalies in arbitrarily distributed highly-dimensional
data.
A soft, or population, anomaly is characterized by a shift in the
distribution of the data set, where certain elements appear with higher
probability than anticipated. Such anomalies must be detected by considering a
sufficiently large sample set rather than a single sample.
Applications include, but not limited to, payment fraud trends, data
exfiltration, disease clusters and epidemics, and social unrests. We evaluate
the method on several domains and obtain both quantitative results and
qualitative insights.
|
Private Sequential Learning | We formulate a private learning model to study an intrinsic tradeoff between
privacy and query complexity in sequential learning. Our model involves a
learner who aims to determine a scalar value, $v^*$, by sequentially querying
an external database and receiving binary responses. In the meantime, an
adversary observes the learner's queries, though not the responses, and tries
to infer from them the value of $v^*$. The objective of the learner is to
obtain an accurate estimate of $v^*$ using only a small number of queries,
while simultaneously protecting her privacy by making $v^*$ provably difficult
to learn for the adversary. Our main results provide tight upper and lower
bounds on the learner's query complexity as a function of desired levels of
privacy and estimation accuracy. We also construct explicit query strategies
whose complexity is optimal up to an additive constant.
|
Automatic Classification of Object Code Using Machine Learning | Recent research has repeatedly shown that machine learning techniques can be
applied to either whole files or file fragments to classify them for analysis.
We build upon these techniques to show that for samples of un-labeled compiled
computer object code, one can apply the same type of analysis to classify
important aspects of the code, such as its target architecture and endianess.
We show that using simple byte-value histograms we retain enough information
about the opcodes within a sample to classify the target architecture with high
accuracy, and then discuss heuristic-based features that exploit information
within the operands to determine endianess. We introduce a dataset with over
16000 code samples from 20 architectures and experimentally show that by using
our features, classifiers can achieve very high accuracy with relatively small
sample sizes.
|
Branching embedding: A heuristic dimensionality reduction algorithm
based on hierarchical clustering | This paper proposes a new dimensionality reduction algorithm named branching
embedding (BE). It converts a dendrogram to a two-dimensional scatter plot, and
visualizes the inherent structures of the original high-dimensional data. Since
the conversion part is not computationally demanding, the BE algorithm would be
beneficial for the case where hierarchical clustering is already performed.
Numerical experiments revealed that the outputs of the algorithm moderately
preserve the original hierarchical structures.
|
Predicting clinical significance of BRCA1 and BRCA2 single nucleotide
substitution variants with unknown clinical significance using probabilistic
neural network and deep neural network-stacked autoencoder | Non-synonymous single nucleotide polymorphisms (nsSNPs) are single nucleotide
substitution occurring in the coding region of a gene and leads to a change in
amino-acid sequence of protein. The studies have shown these variations may be
associated with disease. Thus, investigating the effects of nsSNPs on protein
function will give a greater insight on how nsSNPs can lead into disease.
Breast cancer is the most common cancer among women causing highest cancer
death every year. BRCA1 and BRCA2 tumor suppressor genes are two main
candidates of which, mutations in them can increase the risk of developing
breast cancer. For prediction and detection of the cancer one can use
experimental or computational methods, but the experimental method is very
costly and time consuming in comparison with the computational method. The
computer and computational methods have been used for more than 30 years. Here
we try to predict the clinical significance of BRCA1 and BRCA2 nsSNPs as well
as the unknown clinical significances. Nearly 500 BRCA1 and BRCA2 nsSNPs with
known clinical significances retrieved from NCBI database. Based on
hydrophobicity or hydrophilicity and their role in proteins' second structure,
they are divided into 6 groups, each assigned with scores. The data are
prepared in the acceptable form to the automated prediction mechanisms,
Probabilistic Neural Network (PNN) and Deep Neural NetworkStacked AutoEncoder
(DNN). With Jackknife cross validation we show that the prediction accuracy
achieved for BRCA1 and BRCA2 using PNN are 87.97% and 82.17% respectively,
while 95.41% and 92.80% accuracies achieved using DNN. The total required
processing time for the training and testing the PNN is 0.9 second and DNN
requires about 7 hours of training and it can predict instantly. both methods
show great improvement in accuracy and speed compared to previous attempts.
|
Zero-shot Sequence Labeling: Transferring Knowledge from Sentences to
Tokens | Can attention- or gradient-based visualization techniques be used to infer
token-level labels for binary sequence tagging problems, using networks trained
only on sentence-level labels? We construct a neural network architecture based
on soft attention, train it as a binary sentence classifier and evaluate
against token-level annotation on four different datasets. Inferring token
labels from a network provides a method for quantitatively evaluating what the
model is learning, along with generating useful feedback in assistance systems.
Our results indicate that attention-based methods are able to predict
token-level labels more accurately, compared to gradient-based methods,
sometimes even rivaling the supervised oracle network.
|
Discrete Factorization Machines for Fast Feature-based Recommendation | User and item features of side information are crucial for accurate
recommendation. However, the large number of feature dimensions, e.g., usually
larger than 10^7, results in expensive storage and computational cost. This
prohibits fast recommendation especially on mobile applications where the
computational resource is very limited. In this paper, we develop a generic
feature-based recommendation model, called Discrete Factorization Machine
(DFM), for fast and accurate recommendation. DFM binarizes the real-valued
model parameters (e.g., float32) of every feature embedding into binary codes
(e.g., boolean), and thus supports efficient storage and fast user-item score
computation. To avoid the severe quantization loss of the binarization, we
propose a convergent updating rule that resolves the challenging discrete
optimization of DFM. Through extensive experiments on two real-world datasets,
we show that 1) DFM consistently outperforms state-of-the-art binarized
recommendation models, and 2) DFM shows very competitive performance compared
to its real-valued version (FM), demonstrating the minimized quantization loss.
This work is accepted by IJCAI 2018.
|
Reachability Analysis of Deep Neural Networks with Provable Guarantees | Verifying correctness of deep neural networks (DNNs) is challenging. We study
a generic reachability problem for feed-forward DNNs which, for a given set of
inputs to the network and a Lipschitz-continuous function over its outputs,
computes the lower and upper bound on the function values. Because the network
and the function are Lipschitz continuous, all values in the interval between
the lower and upper bound are reachable. We show how to obtain the safety
verification problem, the output range analysis problem and a robustness
measure by instantiating the reachability problem. We present a novel algorithm
based on adaptive nested optimisation to solve the reachability problem. The
technique has been implemented and evaluated on a range of DNNs, demonstrating
its efficiency, scalability and ability to handle a broader class of networks
than state-of-the-art verification approaches.
|
Bayesian Regularization for Graphical Models with Unequal Shrinkage | We consider a Bayesian framework for estimating a high-dimensional sparse
precision matrix, in which adaptive shrinkage and sparsity are induced by a
mixture of Laplace priors. Besides discussing our formulation from the Bayesian
standpoint, we investigate the MAP (maximum a posteriori) estimator from a
penalized likelihood perspective that gives rise to a new non-convex penalty
approximating the $\ell_0$ penalty. Optimal error rates for estimation
consistency in terms of various matrix norms along with selection consistency
for sparse structure recovery are shown for the unique MAP estimator under mild
conditions. For fast and efficient computation, an EM algorithm is proposed to
compute the MAP estimator of the precision matrix and (approximate) posterior
probabilities on the edges of the underlying sparse structure. Through
extensive simulation studies and a real application to a call center data, we
have demonstrated the fine performance of our method compared with existing
alternatives.
|
Incorporating Privileged Information to Unsupervised Anomaly Detection | We introduce a new unsupervised anomaly detection ensemble called SPI which
can harness privileged information - data available only for training examples
but not for (future) test examples. Our ideas build on the Learning Using
Privileged Information (LUPI) paradigm pioneered by Vapnik et al. [19,17],
which we extend to unsupervised learning and in particular to anomaly
detection. SPI (for Spotting anomalies with Privileged Information) constructs
a number of frames/fragments of knowledge (i.e., density estimates) in the
privileged space and transfers them to the anomaly scoring space through
"imitation" functions that use only the partial information available for test
examples. Our generalization of the LUPI paradigm to unsupervised anomaly
detection shepherds the field in several key directions, including (i) domain
knowledge-augmented detection using expert annotations as PI, (ii) fast
detection using computationally-demanding data as PI, and (iii) early detection
using "historical future" data as PI. Through extensive experiments on
simulated and real datasets, we show that augmenting privileged information to
anomaly detection significantly improves detection performance. We also
demonstrate the promise of SPI under all three settings (i-iii); with PI
capturing expert knowledge, computationally expensive features, and future data
on three real world detection tasks.
|
S4ND: Single-Shot Single-Scale Lung Nodule Detection | The state of the art lung nodule detection studies rely on computationally
expensive multi-stage frameworks to detect nodules from CT scans. To address
this computational challenge and provide better performance, in this paper we
propose S4ND, a new deep learning based method for lung nodule detection. Our
approach uses a single feed forward pass of a single network for detection and
provides better performance when compared to the current literature. The whole
detection pipeline is designed as a single $3D$ Convolutional Neural Network
(CNN) with dense connections, trained in an end-to-end manner. S4ND does not
require any further post-processing or user guidance to refine detection
results. Experimentally, we compared our network with the current
state-of-the-art object detection network (SSD) in computer vision as well as
the state-of-the-art published method for lung nodule detection (3D DCNN). We
used publically available $888$ CT scans from LUNA challenge dataset and showed
that the proposed method outperforms the current literature both in terms of
efficiency and accuracy by achieving an average FROC-score of $0.897$. We also
provide an in-depth analysis of our proposed network to shed light on the
unclear paradigms of tiny object detection.
|
Clustering With Pairwise Relationships: A Generative Approach | Semi-supervised learning (SSL) has become important in current data analysis
applications, where the amount of unlabeled data is growing exponentially and
user input remains limited by logistics and expense. Constrained clustering, as
a subclass of SSL, makes use of user input in the form of relationships between
data points (e.g., pairs of data points belonging to the same class or
different classes) and can remarkably improve the performance of unsupervised
clustering in order to reflect user-defined knowledge of the relationships
between particular data points. Existing algorithms incorporate such user
input, heuristically, as either hard constraints or soft penalties, which are
separate from any generative or statistical aspect of the clustering model;
this results in formulations that are suboptimal and not sufficiently general.
In this paper, we propose a principled, generative approach to
probabilistically model, without ad hoc penalties, the joint distribution given
by user-defined pairwise relations. The proposed model accounts for general
underlying distributions without assuming a specific form and relies on
expectation-maximization for model fitting. For distributions in a standard
form, the proposed approach results in a closed-form solution for updated
parameters.
|
Examining the Use of Neural Networks for Feature Extraction: A
Comparative Analysis using Deep Learning, Support Vector Machines, and
K-Nearest Neighbor Classifiers | Neural networks in many varieties are touted as very powerful machine
learning tools because of their ability to distill large amounts of information
from different forms of data, extracting complex features and enabling powerful
classification abilities. In this study, we use neural networks to extract
features from both images and numeric data and use these extracted features as
inputs for other machine learning models, namely support vector machines (SVMs)
and k-nearest neighbor classifiers (KNNs), in order to see if
neural-network-extracted features enhance the capabilities of these models. We
tested 7 different neural network architectures in this manner, 4 for images
and 3 for numeric data, training each for varying lengths of time and then
comparing the results of the neural network independently to those of an SVM
and KNN on the data, and finally comparing these results to models of SVM and
KNN trained using features extracted via the neural network architecture. This
process was repeated on 3 different image datasets and 2 different numeric
datasets. The results show that, in many cases, the features extracted using
the neural network significantly improve the capabilities of SVMs and KNNs
compared to running these algorithms on the raw features, and in some cases
also surpass the performance of the neural network alone. This in turn suggests
that it may be a reasonable practice to use neural networks as a means to
extract features for classification by other machine learning models for some
datasets.
|
DIRECT: Deep Discriminative Embedding for Clustering of LIGO Data | In this paper, benefiting from the strong ability of deep neural network in
estimating non-linear functions, we propose a discriminative embedding function
to be used as a feature extractor for clustering tasks. The trained embedding
function transfers knowledge from the domain of a labeled set of
morphologically-distinct images, known as classes, to a new domain within which
new classes can potentially be isolated and identified. Our target application
in this paper is the Gravity Spy Project, which is an effort to characterize
transient, non-Gaussian noise present in data from the Advanced Laser
Interferometer Gravitational-wave Observatory, or LIGO. Accumulating large,
labeled sets of noise features and identifying of new classes of noise lead to
a better understanding of their origin, which makes their removal from the data
and/or detectors possible.
|
Semi-orthogonal Non-negative Matrix Factorization with an Application in
Text Mining | Emergency Department (ED) crowding is a worldwide issue that affects the
efficiency of hospital management and the quality of patient care. This occurs
when the request for an admit ward-bed to receive a patient is delayed until an
admission decision is made by a doctor. To reduce the overcrowding and waiting
time of ED, we build a classifier to predict the disposition of patients using
manually-typed nurse notes collected during triage, thereby allowing hospital
staff to begin necessary preparation beforehand. However, these triage notes
involve high dimensional, noisy, and also sparse text data which makes model
fitting and interpretation difficult. To address this issue, we propose the
semi-orthogonal non-negative matrix factorization (SONMF) for both continuous
and binary design matrices to first bi-cluster the patients and words into a
reduced number of topics. The subjects can then be interpreted as a
non-subtractive linear combination of orthogonal basis topic vectors. These
generated topic vectors provide the hospital with a direct understanding of the
cause of admission. We show that by using a transformation of basis, the
classification accuracy can be further increased compared to the conventional
bag-of-words model and alternative matrix factorization approaches. Through
simulated data experiments, we also demonstrate that the proposed method
outperforms other non-negative matrix factorization (NMF) methods in terms of
factorization accuracy, rate of convergence, and degree of orthogonality.
|
Implementation of Stochastic Quasi-Newton's Method in PyTorch | In this paper, we implement the Stochastic Damped LBFGS (SdLBFGS) for
stochastic non-convex optimization. We make two important modifications to the
original SdLBFGS algorithm. First, by initializing the Hessian at each step
using an identity matrix, the algorithm converges better than original
algorithm. Second, by performing direction normalization we could gain stable
optimization procedure without line search. Experiments on minimizing a 2D
non-convex function shows that our improved algorithm converges better than
original algorithm, and experiments on the CIFAR10 and MNIST datasets show that
our improved algorithm works stably and gives comparable or even better testing
accuracies than first order optimizers SGD, Adagrad, and second order
optimizers LBFGS in PyTorch.
|
(Nearly) Efficient Algorithms for the Graph Matching Problem on
Correlated Random Graphs | We give a quasipolynomial time algorithm for the graph matching problem (also
known as noisy or robust graph isomorphism) on correlated random graphs.
Specifically, for every $\gamma>0$, we give a $n^{O(\log n)}$ time algorithm
that given a pair of $\gamma$-correlated $G(n,p)$ graphs $G_0,G_1$ with average
degree between $n^{\varepsilon}$ and $n^{1/153}$ for $\varepsilon = o(1)$,
recovers the "ground truth" permutation $\pi\in S_n$ that matches the vertices
of $G_0$ to the vertices of $G_n$ in the way that minimizes the number of
mismatched edges. We also give a recovery algorithm for a denser regime, and a
polynomial-time algorithm for distinguishing between correlated and
uncorrelated graphs.
Prior work showed that recovery is information-theoretically possible in this
model as long the average degree was at least $\log n$, but sub-exponential
time algorithms were only known in the dense case (i.e., for $p > n^{-o(1)}$).
Moreover, "Percolation Graph Matching", which is the most common heuristic for
this problem, has been shown to require knowledge of $n^{\Omega(1)}$ "seeds"
(i.e., input/output pairs of the permutation $\pi$) to succeed in this regime.
In contrast our algorithms require no seed and succeed for $p$ which is as low
as $n^{o(1)-1}$.
|
Efficient active learning of sparse halfspaces | We study the problem of efficient PAC active learning of homogeneous linear
classifiers (halfspaces) in $\mathbb{R}^d$, where the goal is to learn a
halfspace with low error using as few label queries as possible. Under the
extra assumption that there is a $t$-sparse halfspace that performs well on the
data ($t \ll d$), we would like our active learning algorithm to be {\em
attribute efficient}, i.e. to have label requirements sublinear in $d$. In this
paper, we provide a computationally efficient algorithm that achieves this
goal. Under certain distributional assumptions on the data, our algorithm
achieves a label complexity of $O(t \cdot \mathrm{polylog}(d, \frac 1
\epsilon))$. In contrast, existing algorithms in this setting are either
computationally inefficient, or subject to label requirements polynomial in $d$
or $\frac 1 \epsilon$.
|
Billion-scale Network Embedding with Iterative Random Projection | Network embedding, which learns low-dimensional vector representation for
nodes in the network, has attracted considerable research attention recently.
However, the existing methods are incapable of handling billion-scale networks,
because they are computationally expensive and, at the same time, difficult to
be accelerated by distributed computing schemes. To address these problems, we
propose RandNE (Iterative Random Projection Network Embedding), a novel and
simple billion-scale network embedding method. Specifically, we propose a
Gaussian random projection approach to map the network into a low-dimensional
embedding space while preserving the high-order proximities between nodes. To
reduce the time complexity, we design an iterative projection procedure to
avoid the explicit calculation of the high-order proximities. Theoretical
analysis shows that our method is extremely efficient, and friendly to
distributed computing schemes without any communication cost in the
calculation. We also design a dynamic updating procedure which can efficiently
incorporate the dynamic changes of the networks without error aggregation.
Extensive experimental results demonstrate the efficiency and efficacy of
RandNE over state-of-the-art methods in several tasks including network
reconstruction, link prediction and node classification on multiple datasets
with different scales, ranging from thousands to billions of nodes and edges.
|
Relating Eye-Tracking Measures With Changes In Knowledge on Search Tasks | We conducted an eye-tracking study where 30 participants performed searches
on the web. We measured their topical knowledge before and after each task.
Their eye-fixations were labelled as "reading" or "scanning". The series of
reading fixations in a line, called "reading-sequences" were characterized by
their length in pixels, fixation duration, and the number of fixations making
up the sequence. We hypothesize that differences in knowledge-change of
participants are reflected in their eye-tracking measures related to reading.
Our results show that the participants with higher change in knowledge differ
significantly in terms of their total reading-sequence-length,
reading-sequence-duration, and number of reading fixations, when compared to
participants with lower knowledge-change.
|
Prioritizing network communities | Uncovering modular structure in networks is fundamental for systems in
biology, physics, and engineering. Community detection identifies candidate
modules as hypotheses, which then need to be validated through experiments,
such as mutagenesis in a biological laboratory. Only a few communities can
typically be validated, and it is thus important to prioritize which
communities to select for downstream experimentation. Here we develop CRank, a
mathematically principled approach for prioritizing network communities. CRank
efficiently evaluates robustness and magnitude of structural features of each
community and then combines these features into the community prioritization.
CRank can be used with any community detection method. It needs only
information provided by the network structure and does not require any
additional metadata or labels. However, when available, CRank can incorporate
domain-specific information to further boost performance. Experiments on many
large networks show that CRank effectively prioritizes communities, yielding a
nearly 50-fold improvement in community prioritization.
|
Sentence-State LSTM for Text Representation | Bi-directional LSTMs are a powerful tool for text representation. On the
other hand, they have been shown to suffer various limitations due to their
sequential nature. We investigate an alternative LSTM structure for encoding
text, which consists of a parallel state for each word. Recurrent steps are
used to perform local and global information exchange between words
simultaneously, rather than incremental reading of a sequence of words. Results
on various classification and sequence labelling benchmarks show that the
proposed model has strong representation power, giving highly competitive
performances compared to stacked BiLSTM models with similar parameter numbers.
|
The Logistic Network Lasso | We apply the network Lasso to solve binary classification and clustering
problems for network-structured data. To this end, we generalize ordinary
logistic regression to non-Euclidean data with an intrinsic network structure.
The resulting "logistic network Lasso" amounts to solving a non-smooth convex
regularized empirical risk minimization. The risk is measured using the
logistic loss incurred over a small set of labeled nodes. For the
regularization, we propose to use the total variation of the classifier
requiring it to conform to the underlying network structure. A scalable
implementation of the learning method is obtained using an inexact variant of
the alternating direction methods of multipliers which results in a scalable
learning algorithm
|
Transformer for Emotion Recognition | This paper describes the UMONS solution for the OMG-Emotion Challenge. We
explore a context-dependent architecture where the arousal and valence of an
utterance are predicted according to its surrounding context (i.e. the
preceding and following utterances of the video). We report an improvement when
taking into account context for both unimodal and multimodal predictions.
|
A probabilistic framework for handwritten text line segmentation | We successfully combine Expectation-Maximization algorithm and variational
approaches for parameter learning and computing inference on Markov random
felds. This is a general method that can be applied to many computer vision
tasks. In this paper, we apply it to handwritten text line segmentation. We
conduct several experiments that demonstrate that our method deal with common
issues of this task, such as complex document layout or non-latin scripts. The
obtained results prove that our method achieve state-of-the-art performance on
different benchmark datasets without any particular fine tuning step.
|
Understanding Reuse, Performance, and Hardware Cost of DNN Dataflows: A
Data-Centric Approach Using MAESTRO | The data partitioning and scheduling strategies used by DNN accelerators to
leverage reuse and perform staging are known as dataflow, and they directly
impact the performance and energy efficiency of DNN accelerator designs. An
accelerator microarchitecture dictates the dataflow(s) that can be employed to
execute a layer or network. Selecting an optimal dataflow for a layer shape can
have a large impact on utilization and energy efficiency, but there is a lack
of understanding on the choices and consequences of dataflows, and of tools and
methodologies to help architects explore the co-optimization design space. In
this work, we first introduce a set of data-centric directives to concisely
specify the space of DNN dataflows in a compilerfriendly form. We then show how
these directives can be analyzed to infer various forms of reuse and to exploit
them using hardware capabilities. We codify this analysis into an analytical
cost model, MAESTRO (Modeling Accelerator Efficiency via Spatio-Temporal Reuse
and Occupancy), that estimates various cost-benefit tradeoffs of a dataflow
including execution time and energy efficiency for a DNN model and hardware
configuration. We demonstrate the use of MAESTRO to drive a hardware design
space exploration (DSE) experiment, which searches across 480M designs to
identify 2.5M valid designs at an average rate of 0.17M designs per second,
including Pareto-optimal throughput- and energy-optimized design points.
|
Sharp Analysis of a Simple Model for Random Forests | Random forests have become an important tool for improving accuracy in
regression and classification problems since their inception by Leo Breiman in
2001. In this paper, we revisit a historically important random forest model
originally proposed by Breiman in 2004 and later studied by G\'erard Biau in
2012, where a feature is selected at random and the splits occurs at the
midpoint of the node along the chosen feature. If the regression function is
Lipschitz and depends only on a small subset of $ S $ out of $ d $ features, we
show that, given access to $ n $ observations and properly tuned split
probabilities, the mean-squared prediction error is $ O((n(\log
n)^{(S-1)/2})^{-\frac{1}{S\log2+1}}) $. This positively answers an outstanding
question of Biau about whether the rate of convergence for this random forest
model could be improved. Furthermore, by a refined analysis of the
approximation and estimation errors for linear models, we show that this rate
cannot be improved in general. Finally, we generalize our analysis and improve
extant prediction error bounds for another random forest model in which each
tree is constructed from subsampled data and the splits are performed at the
empirical median along a chosen feature.
|
Modeling Dengue Vector Population Using Remotely Sensed Data and Machine
Learning | Mosquitoes are vectors of many human diseases. In particular, Aedes \ae gypti
(Linnaeus) is the main vector for Chikungunya, Dengue, and Zika viruses in
Latin America and it represents a global threat. Public health policies that
aim at combating this vector require dependable and timely information, which
is usually expensive to obtain with field campaigns. For this reason, several
efforts have been done to use remote sensing due to its reduced cost. The
present work includes the temporal modeling of the oviposition activity
(measured weekly on 50 ovitraps in a north Argentinean city) of Aedes \ae gypti
(Linnaeus), based on time series of data extracted from operational earth
observation satellite images. We use are NDVI, NDWI, LST night, LST day and
TRMM-GPM rain from 2012 to 2016 as predictive variables. In contrast to
previous works which use linear models, we employ Machine Learning techniques
using completely accessible open source toolkits. These models have the
advantages of being non-parametric and capable of describing nonlinear
relationships between variables. Specifically, in addition to two linear
approaches, we assess a Support Vector Machine, an Artificial Neural Networks,
a K-nearest neighbors and a Decision Tree Regressor. Considerations are made on
parameter tuning and the validation and training approach. The results are
compared to linear models used in previous works with similar data sets for
generating temporal predictive models. These new tools perform better than
linear approaches, in particular Nearest Neighbor Regression (KNNR) performs
the best. These results provide better alternatives to be implemented
operatively on the Argentine geospatial Risk system that is running since 2012.
|
Computing the Shattering Coefficient of Supervised Learning Algorithms | The Statistical Learning Theory (SLT) provides the theoretical guarantees for
supervised machine learning based on the Empirical Risk Minimization Principle
(ERMP). Such principle defines an upper bound to ensure the uniform convergence
of the empirical risk Remp(f), i.e., the error measured on a given data sample,
to the expected value of risk R(f) (a.k.a. actual risk), which depends on the
Joint Probability Distribution P(X x Y) mapping input examples x in X to class
labels y in Y. The uniform convergence is only ensured when the Shattering
coefficient N(F,2n) has a polynomial growing behavior. This paper proves the
Shattering coefficient for any Hilbert space H containing the input space X and
discusses its effects in terms of learning guarantees for supervised machine
algorithms.
|
Wavelet Decomposition of Gradient Boosting | In this paper we introduce a significant improvement to the popular
tree-based Stochastic Gradient Boosting algorithm using a wavelet decomposition
of the trees. This approach is based on harmonic analysis and approximation
theoretical elements, and as we show through extensive experimentation, our
wavelet based method generally outperforms existing methods, particularly in
difficult scenarios of class unbalance and mislabeling in the training data.
|
Gradient Descent for One-Hidden-Layer Neural Networks: Polynomial
Convergence and SQ Lower Bounds | We study the complexity of training neural network models with one hidden
nonlinear activation layer and an output weighted sum layer. We analyze
Gradient Descent applied to learning a bounded target function on $n$
real-valued inputs. We give an agnostic learning guarantee for GD: starting
from a randomly initialized network, it converges in mean squared loss to the
minimum error (in $2$-norm) of the best approximation of the target function
using a polynomial of degree at most $k$. Moreover, for any $k$, the size of
the network and number of iterations needed are both bounded by
$n^{O(k)}\log(1/\epsilon)$. In particular, this applies to training networks of
unbiased sigmoids and ReLUs. We also rigorously explain the empirical finding
that gradient descent discovers lower frequency Fourier components before
higher frequency components.
We complement this result with nearly matching lower bounds in the
Statistical Query model. GD fits well in the SQ framework since each training
step is determined by an expectation over the input distribution. We show that
any SQ algorithm that achieves significant improvement over a constant function
with queries of tolerance some inverse polynomial in the input dimensionality
$n$ must use $n^{\Omega(k)}$ queries even when the target functions are
restricted to a set of $n^{O(k)}$ degree-$k$ polynomials, and the input
distribution is uniform over the unit sphere; for this class the
information-theoretic lower bound is only $\Theta(k \log n)$.
Our approach for both parts is based on spherical harmonics. We view gradient
descent as an operator on the space of functions, and study its dynamics. An
essential tool is the Funk-Hecke theorem, which explains the eigenfunctions of
this operator in the case of the mean squared loss.
|
Holarchic Structures for Decentralized Deep Learning - A Performance
Analysis | Structure plays a key role in learning performance. In centralized
computational systems, hyperparameter optimization and regularization
techniques such as dropout are computational means to enhance learning
performance by adjusting the deep hierarchical structure. However, in
decentralized deep learning by the Internet of Things, the structure is an
actual network of autonomous interconnected devices such as smart phones that
interact via complex network protocols. Self-adaptation of the learning
structure is a challenge. Uncertainties such as network latency, node and link
failures or even bottlenecks by limited processing capacity and energy
availability can signif- icantly downgrade learning performance. Network
self-organization and self-management is complex, while it requires additional
computational and network resources that hinder the feasibility of
decentralized deep learning. In contrast, this paper introduces a self-adaptive
learning approach based on holarchic learning structures for exploring,
mitigating and boosting learning performance in distributed environments with
uncertainties. A large-scale performance analysis with 864000 experiments fed
with synthetic and real-world data from smart grid and smart city pilot
projects confirm the cost-effectiveness of holarchic structures for
decentralized deep learning.
|
Real-time regression analysis with deep convolutional neural networks | We discuss the development of novel deep learning algorithms to enable
real-time regression analysis for time series data. We showcase the application
of this new method with a timely case study, and then discuss the applicability
of this approach to tackle similar challenges across science domains.
|
Detecting Compressed Cleartext Traffic from Consumer Internet of Things
Devices | Data encryption is the primary method of protecting the privacy of consumer
device Internet communications from network observers. The ability to
automatically detect unencrypted data in network traffic is therefore an
essential tool for auditing Internet-connected devices. Existing methods
identify network packets containing cleartext but cannot differentiate packets
containing encrypted data from packets containing compressed unencrypted data,
which can be easily recovered by reversing the compression algorithm. This
makes it difficult for consumer protection advocates to identify devices that
risk user privacy by sending sensitive data in a compressed unencrypted format.
Here, we present the first technique to automatically distinguish encrypted
from compressed unencrypted network transmissions on a per-packet basis. We
apply three machine learning models and achieve a maximum 66.9% accuracy with a
convolutional neural network trained on raw packet data. This result is a
baseline for this previously unstudied machine learning problem, which we hope
will motivate further attention and accuracy improvements. To facilitate
continuing research on this topic, we have made our training and test datasets
available to the public.
|
Building Disease Detection Algorithms with Very Small Numbers of
Positive Samples | Although deep learning can provide promising results in medical image
analysis, the lack of very large annotated datasets confines its full
potential. Furthermore, limited positive samples also create unbalanced
datasets which limit the true positive rates of trained models. As unbalanced
datasets are mostly unavoidable, it is greatly beneficial if we can extract
useful knowledge from negative samples to improve classification accuracy on
limited positive samples. To this end, we propose a new strategy for building
medical image analysis pipelines that target disease detection. We train a
discriminative segmentation model only on normal images to provide a source of
knowledge to be transferred to a disease detection classifier. We show that
using the feature maps of a trained segmentation network, deviations from
normal anatomy can be learned by a two-class classification network on an
extremely unbalanced training dataset with as little as one positive for 17
negative samples. We demonstrate that even though the segmentation network is
only trained on normal cardiac computed tomography images, the resulting
feature maps can be used to detect pericardial effusion and cardiac septal
defects with two-class convolutional classification networks.
|
What game are we playing? End-to-end learning in normal and extensive
form games | Although recent work in AI has made great progress in solving large,
zero-sum, extensive-form games, the underlying assumption in most past work is
that the parameters of the game itself are known to the agents. This paper
deals with the relatively under-explored but equally important "inverse"
setting, where the parameters of the underlying game are not known to all
agents, but must be learned through observations. We propose a differentiable,
end-to-end learning framework for addressing this task. In particular, we
consider a regularized version of the game, equivalent to a particular form of
quantal response equilibrium, and develop 1) a primal-dual Newton method for
finding such equilibrium points in both normal and extensive form games; and 2)
a backpropagation method that lets us analytically compute gradients of all
relevant game parameters through the solution itself. This ultimately lets us
learn the game by training in an end-to-end fashion, effectively by integrating
a "differentiable game solver" into the loop of larger deep network
architectures. We demonstrate the effectiveness of the learning method in
several settings including poker and security game tasks.
|
Fast Online Exact Solutions for Deterministic MDPs with Sparse Rewards | Markov Decision Processes (MDPs) are a mathematical framework for modeling
sequential decision making under uncertainty. The classical approaches for
solving MDPs are well known and have been widely studied, some of which rely on
approximation techniques to solve MDPs with large state space and/or action
space. However, most of these classical solution approaches and their
approximation techniques still take much computation time to converge and
usually must be re-computed if the reward function is changed. This paper
introduces a novel alternative approach for exactly and efficiently solving
deterministic, continuous MDPs with sparse reward sources. When the environment
is such that the "distance" between states can be determined in constant time,
e.g. grid world, our algorithm offers $O( |R|^2 \times |A|^2 \times |S|)$,
where $|R|$ is the number of reward sources, $|A|$ is the number of actions,
and $|S|$ is the number of states. Memory complexity for the algorithm is $O(
|S| + |R| \times |A|)$. This new approach opens new avenues for boosting
computational performance for certain classes of MDPs and is of tremendous
value for MDP applications such as robotics and unmanned systems. This paper
describes the algorithm and presents numerical experiment results to
demonstrate its powerful computational performance. We also provide rigorous
mathematical description of the approach.
|
ReGAN: RE[LAX|BAR|INFORCE] based Sequence Generation using GANs | Generative Adversarial Networks (GANs) have seen steep ascension to the peak
of ML research zeitgeist in recent years. Mostly catalyzed by its success in
the domain of image generation, the technique has seen wide range of adoption
in a variety of other problem domains. Although GANs have had a lot of success
in producing more realistic images than other approaches, they have only seen
limited use for text sequences. Generation of longer sequences compounds this
problem. Most recently, SeqGAN (Yu et al., 2017) has shown improvements in
adversarial evaluation and results with human evaluation compared to a MLE
based trained baseline. The main contributions of this paper are three-fold: 1.
We show results for sequence generation using a GAN architecture with efficient
policy gradient estimators, 2. We attain improved training stability, and 3. We
perform a comparative study of recent unbiased low variance gradient estimation
techniques such as REBAR (Tucker et al., 2017), RELAX (Grathwohl et al., 2018)
and REINFORCE (Williams, 1992). Using a simple grammar on synthetic datasets
with varying length, we indicate the quality of sequences generated by the
model.
|
Several Tunable GMM Kernels | While tree methods have been popular in practice, researchers and
practitioners are also looking for simple algorithms which can reach similar
accuracy of trees. In 2010, (Ping Li UAI'10) developed the method of
"abc-robust-logitboost" and compared it with other supervised learning methods
on datasets used by the deep learning literature. In this study, we propose a
series of "tunable GMM kernels" which are simple and perform largely comparably
to tree methods on the same datasets. Note that "abc-robust-logitboost"
substantially improved the original "GDBT" in that (a) it developed a
tree-split formula based on second-order information of the derivatives of the
loss function; (b) it developed a new set of derivatives for multi-class
classification formulation.
In the prior study in 2017, the "generalized min-max" (GMM) kernel was shown
to have good performance compared to the "radial-basis function" (RBF) kernel.
However, as demonstrated in this paper, the original GMM kernel is often not as
competitive as tree methods on the datasets used in the deep learning
literature. Since the original GMM kernel has no parameters, we propose tunable
GMM kernels by adding tuning parameters in various ways. Three basic (i.e.,
with only one parameter) GMM kernels are the "$e$GMM kernel", "$p$GMM kernel",
and "$\gamma$GMM kernel", respectively. Extensive experiments show that they
are able to produce good results for a large number of classification tasks.
Furthermore, the basic kernels can be combined to boost the performance.
|
Fighting Accounting Fraud Through Forensic Data Analytics | Accounting fraud is a global concern representing a significant threat to the
financial system stability due to the resulting diminishing of the market
confidence and trust of regulatory authorities. Several tricks can be used to
commit accounting fraud, hence the need for non-static regulatory interventions
that take into account different fraudulent patterns. Accordingly, this study
aims to improve the detection of accounting fraud via the implementation of
several machine learning methods to better differentiate between fraud and
non-fraud companies, and to further assist the task of examination within the
riskier firms by evaluating relevant financial indicators. Out-of-sample
results suggest there is a great potential in detecting falsified financial
statements through statistical modelling and analysis of publicly available
accounting information. The proposed methodology can be of assistance to public
auditors and regulatory agencies as it facilitates auditing processes, and
supports more targeted and effective examinations of accounting reports.
|
Identifiability of Generalized Hypergeometric Distribution (GHD)
Directed Acyclic Graphical Models | We introduce a new class of identifiable DAG models where the conditional
distribution of each node given its parents belongs to a family of generalized
hypergeometric distributions (GHD). A family of generalized hypergeometric
distributions includes a lot of discrete distributions such as the binomial,
Beta-binomial, negative binomial, Poisson, hyper-Poisson, and many more. We
prove that if the data drawn from the new class of DAG models, one can fully
identify the graph structure. We further present a reliable and polynomial-time
algorithm that recovers the graph from finitely many data. We show through
theoretical results and numerical experiments that our algorithm is
statistically consistent in high-dimensional settings (p>n) if the indegree of
the graph is bounded, and out-performs state-of-the-art DAG learning
algorithms.
|
Tile2Vec: Unsupervised representation learning for spatially distributed
data | Geospatial analysis lacks methods like the word vector representations and
pre-trained networks that significantly boost performance across a wide range
of natural language and computer vision tasks. To fill this gap, we introduce
Tile2Vec, an unsupervised representation learning algorithm that extends the
distributional hypothesis from natural language -- words appearing in similar
contexts tend to have similar meanings -- to spatially distributed data. We
demonstrate empirically that Tile2Vec learns semantically meaningful
representations on three datasets. Our learned representations significantly
improve performance in downstream classification tasks and, similar to word
vectors, visual analogies can be obtained via simple arithmetic in the latent
space.
|
Survey and cross-benchmark comparison of remaining time prediction
methods in business process monitoring | Predictive business process monitoring methods exploit historical process
execution logs to generate predictions about running instances (called cases)
of a business process, such as the prediction of the outcome, next activity or
remaining cycle time of a given process case. These insights could be used to
support operational managers in taking remedial actions as business processes
unfold, e.g. shifting resources from one case onto another to ensure this
latter is completed on time. A number of methods to tackle the remaining cycle
time prediction problem have been proposed in the literature. However, due to
differences in their experimental setup, choice of datasets, evaluation
measures and baselines, the relative merits of each method remain unclear. This
article presents a systematic literature review and taxonomy of methods for
remaining time prediction in the context of business processes, as well as a
cross-benchmark comparison of 16 such methods based on 16 real-life datasets
originating from different industry domains.
|
Profitable Bandits | Originally motivated by default risk management applications, this paper
investigates a novel problem, referred to as the profitable bandit problem
here. At each step, an agent chooses a subset of the K possible actions. For
each action chosen, she then receives the sum of a random number of rewards.
Her objective is to maximize her cumulated earnings. We adapt and study three
well-known strategies in this purpose, that were proved to be most efficient in
other settings: kl-UCB, Bayes-UCB and Thompson Sampling. For each of them, we
prove a finite time regret bound which, together with a lower bound we obtain
as well, establishes asymptotic optimality. Our goal is also to compare these
three strategies from a theoretical and empirical perspective both at the same
time. We give simple, self-contained proofs that emphasize their similarities,
as well as their differences. While both Bayesian strategies are automatically
adapted to the geometry of information, the numerical experiments carried out
show a slight advantage for Thompson Sampling in practice.
|
Interpretable Adversarial Perturbation in Input Embedding Space for Text | Following great success in the image processing field, the idea of
adversarial training has been applied to tasks in the natural language
processing (NLP) field. One promising approach directly applies adversarial
training developed in the image processing field to the input word embedding
space instead of the discrete input space of texts. However, this approach
abandons such interpretability as generating adversarial texts to significantly
improve the performance of NLP tasks. This paper restores interpretability to
such methods by restricting the directions of perturbations toward the existing
words in the input embedding space. As a result, we can straightforwardly
reconstruct each input with perturbations to an actual text by considering the
perturbations to be the replacement of words in the sentence while maintaining
or even improving the task performance.
|
Multinomial Logit Bandit with Linear Utility Functions | Multinomial logit bandit is a sequential subset selection problem which
arises in many applications. In each round, the player selects a
$K$-cardinality subset from $N$ candidate items, and receives a reward which is
governed by a {\it multinomial logit} (MNL) choice model considering both item
utility and substitution property among items. The player's objective is to
dynamically learn the parameters of MNL model and maximize cumulative reward
over a finite horizon $T$. This problem faces the exploration-exploitation
dilemma, and the involved combinatorial nature makes it non-trivial. In recent
years, there have developed some algorithms by exploiting specific
characteristics of the MNL model, but all of them estimate the parameters of
MNL model separately and incur a regret no better than
$\tilde{O}\big(\sqrt{NT}\big)$ which is not preferred for large candidate set
size $N$. In this paper, we consider the {\it linear utility} MNL choice model
whose item utilities are represented as linear functions of $d$-dimension item
features, and propose an algorithm, titled {\bf LUMB}, to exploit the
underlying structure. It is proven that the proposed algorithm achieves
$\tilde{O}\big(dK\sqrt{T}\big)$ regret which is free of candidate set size.
Experiments show the superiority of the proposed algorithm.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.