IT博客汇
  • 首页
  • 精华
  • 技术
  • 设计
  • 资讯
  • 扯淡
  • 权利声明
  • 登录 注册

    arXiv Paper Daily: Tue, 3 Jan 2017

    我爱机器学习(52ml.net)发表于 2017-01-03 00:00:00
    love 0

    Computer Vision and Pattern Recognition

    Retrieving Similar X-Ray Images from Big Image Data Using Radon Barcodes with Single Projections

    Morteza Babaie, H.R. Tizhoosh, Shujin Zhu, M.E. Shiri
    Comments: Accepted for publication in ICPRAM 2017: The International Conference on Pattern Recognition Applications and Methods, Porto, Portugal, 2017
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    The idea of Radon barcodes (RBC) has been introduced recently. In this paper,
    we propose a content-based image retrieval approach for big datasets based on
    Radon barcodes. Our method (Single Projection Radon Barcode, or SP-RBC) uses
    only a few Radon single projections for each image as global features that can
    serve as a basis for weak learners. This is our most important contribution in
    this work, which improves the results of the RBC considerably. As a matter of
    fact, only one projection of an image, as short as a single SURF feature
    vector, can already achieve acceptable results. Nevertheless, using multiple
    projections in a long vector will not deliver anticipated improvements. To
    exploit the information inherent in each projection, our method uses the
    outcome of each projection separately and then applies more precise local
    search on the small subset of retrieved images. We have tested our method using
    IRMA 2009 dataset a with 14,400 x-ray images as part of imageCLEF initiative.
    Our approach leads to a substantial decrease in the error rate in comparison
    with other non-learning methods.

    Adversarially Tuned Scene Generation

    V S R Veeravasarapu, Constantin Rothkopf, Ramesh Visvanathan
    Comments: 9 pages
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Generalization performance of trained computer vision systems that use
    computer graphics (CG) generated data is not yet effective due to the concept
    of ‘domain-shift’ between virtual and real data. Although simulated data
    augmented with a few real world samples has been shown to mitigate domain shift
    and improve transferability of trained models, guiding or bootstrapping the
    virtual data generation with the distributions learnt from target real world
    domain is desired, especially in the fields where annotating even few real
    images is laborious (such as semantic labeling, and intrinsic images etc.). In
    order to address this problem in an unsupervised manner, our work combines
    recent advances in CG (which aims to generate stochastic scene layouts coupled
    with large collections of 3D object models) and generative adversarial training
    (which aims train generative models by measuring discrepancy between generated
    and real data in terms of their separability in the space of a deep
    discriminatively-trained classifier). Our method uses iterative estimation of
    the posterior density of prior distributions for a generative graphical model.
    This is done within a rejection sampling framework. Initially, we assume
    uniform distributions as priors on the parameters of a scene described by a
    generative graphical model. As iterations proceed the prior distributions get
    updated to distributions that are closer to the (unknown) distributions of
    target data. We demonstrate the utility of adversarially tuned scene generation
    on two real-world benchmark datasets (CityScapes and CamVid) for traffic scene
    semantic labeling with a deep convolutional net (DeepLab). We realized
    performance improvements by 2.28 and 3.14 points (using the IoU metric) between
    the DeepLab models trained on simulated sets prepared from the scene generation
    models before and after tuning to CityScapes and CamVid respectively.

    Weakly Supervised Semantic Segmentation using Web-Crawled Videos

    Seunghoon Hong, Donghun Yeo, Suha Kwak, Honglak Lee, Bohyung Han
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    We propose a novel algorithm for weakly supervised semantic segmentation
    based on image-level class labels only. In weakly supervised setting, it is
    commonly observed that trained model overly focuses on discriminative parts
    rather than the entire object area. Our goal is to overcome this limitation
    with no additional human intervention by retrieving videos relevant to target
    class labels from web repository, and generating segmentation labels from the
    retrieved videos to simulate strong supervision for semantic segmentation.
    During this process, we take advantage of image classification with
    discriminative localization technique to reject false alarms in retrieved
    videos and identify relevant spatio-temporal volumes within retrieved videos.
    Although the entire procedure does not require any additional supervision, the
    segmentation annotations obtained from videos are sufficiently strong to learn
    a model for semantic segmentation. The proposed algorithm substantially
    outperforms existing methods based on the same level of supervision and is even
    as competitive as the approaches relying on extra annotations.

    Challenges ahead Electron Microscopy for Structural Biology from the Image Processing point of view

    Carlos Oscar S. Sorzano, Jose Maria Carazo
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Quantitative Methods (q-bio.QM)

    Since the introduction of Direct Electron Detectors (DEDs), the resolution
    and range of macromolecules amenable to this technique has significantly
    widened, generating a broad interest that explains the well over a dozen
    reviews in top journal in the last two years. Similarly, the number of job
    offers to lead EM groups and/or coordinate EM facilities has exploded, and FEI
    (the main microscope manufacturer for Life Sciences) has received more than 100
    orders of high-end electron microscopes by summer 2016. Strategic corporate
    movements are also happening, with very big players entering the market through
    key acquisitions (Thermo Fisher has recently bought FEI for \(4.2B), partly
    attracted by new Pharma interest in the field, now perceived to be in a
    position to impact structure-based drug design. The scientific perspectives are
    indeed extremely positive but, in these moments of well-founded generalized
    optimists, we want to make a reflection on some of the hurdles ahead us, since
    they certainly exist and they indeed limit the informational content of cryoEM
    projects. Here we focus on image processing aspects, particularly in the
    so-called area of Single Particle Analysis, discussing some of the current
    resolution and high-throughput limiting factors.

    Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image

    Denis Tome, Chris Russell, Lourdes Agapito
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    We propose a unified formulation for the problem of 3D human pose estimation
    from a single raw RGB image that reasons jointly about 2D joint estimation and
    3D pose reconstruction to improve both tasks. We take an integrated approach
    that fuses probabilistic knowledge of 3D human pose with a multi-stage CNN
    architecture and uses the knowledge of plausible 3D landmark locations to
    refine the search for better 2D locations. The entire process is trained
    end-to-end, is extremely efficient and obtains state- of-the-art results on
    Human3.6M outperforming previous approaches both on 2D and 3D errors.

    The Geodesic Distance between )mathcal{G}_I^0( Models and its Application to Region Discrimination

    José Naranjo-Torres, Juliana Gambini, Alejandro C. Frery
    Comments: Accepted for publication in the IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (J-STARS), 1 January 2017
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)

    The )mathcal{G}_I^0( distribution is able to characterize different regions
    in monopolarized SAR imagery. It is indexed by three parameters: the number of
    looks (which can be estimated in the whole image), a scale parameter and a
    texture parameter. This paper presents a new proposal for feature extraction
    and region discrimination in SAR imagery, using the geodesic distance as a
    measure of dissimilarity between )mathcal{G}_I^0( models. We derive geodesic
    distances between models that describe several practical situations, assuming
    the number of looks known, for same and different texture and for same and
    different scale. We then apply this new tool to the problems of (i)~identifying
    edges between regions with different texture, and (ii)~quantify the
    dissimilarity between pairs of samples in actual SAR data. We analyze the
    advantages of using the geodesic distance when compared to stochastic
    distances.

    A robust approach for tree segmentation in deciduous forests using small-footprint airborne LiDAR data

    Hamid Hamraz, Marco A. Contreras, Jun Zhang
    Journal-ref: International Journal of Applied Earth Observation and
    Geoinformation 52 (pp. 532-541): Elsevier (2016)
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Computational Engineering, Finance, and Science (cs.CE); Computational Geometry (cs.CG)

    This paper presents a non-parametric approach for segmenting trees from
    airborne LiDAR data in deciduous forests. Based on the LiDAR point cloud, the
    approach collects crown information such as steepness and height on-the-fly to
    delineate crown boundaries, and most importantly, does not require a priori
    assumptions of crown shape and size. The approach segments trees iteratively
    starting from the tallest within a given area to the smallest until all trees
    have been segmented. To evaluate its performance, the approach was applied to
    the University of Kentucky Robinson Forest, a deciduous closed-canopy forest
    with complex terrain and vegetation conditions. The approach identified 94% of
    dominant and co-dominant trees with a false detection rate of 13%. About 62% of
    intermediate, overtopped, and dead trees were also detected with a false
    detection rate of 15%. The overall segmentation accuracy was 77%. Correlations
    of the segmentation scores of the proposed approach with local terrain and
    stand metrics was not significant, which is likely an indication of the
    robustness of the approach as results are not sensitive to the differences in
    terrain and stand structures.

    Video-based Person Re-identification with Accumulative Motion Context

    Hao Liu, Zequn Jie, Karlekar Jayashree, Meibin Qi, Jianguo Jiang, Shuicheng Yan, Jiashi Feng
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Video based person re-identification plays a central role in realistic
    security and video surveillance. In this paper we propose a novel Accumulative
    Motion Context (AMOC) network for addressing this important problem, which
    effectively exploits the long-range motion context for robustly identifying the
    same person under challenging conditions. Given a video sequence of the same or
    different persons, the proposed AMOC network jointly learns appearance
    representation and motion context from a collection of adjacent frames using a
    two-stream convolutional architecture. Then AMOC accumulates clues from motion
    context by recurrent aggregation, allowing effective information flow among
    adjacent frames and capturing dynamic gist of the persons. The architecture of
    AMOC is end-to-end trainable and thus motion context can be adapted to
    complement appearance clues under unfavorable conditions ( extit{e.g.},
    occlusions). Extensive experiments are conduced on two public benchmark
    datasets, extit{i.e.}, the iLIDS-VID and PRID-2011 datasets, to investigate
    the performance of AMOC. The experimental results demonstrate that the proposed
    AMOC network outperforms state-of-the-arts for video-based re-identification
    significantly and confirm the advantage of exploiting long-range motion context
    for video based person re-identification, validating our motivation evidently.

    Tree segmentation in multi-story stands within small-footprint airborne LiDAR data

    Hamid Hamraz, Marco A. Contreras, Jun Zhang
    Comments: 28 double-spaced pages including 6 figures, 7 tables, and 50 references. The manuscript is currently under review
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Computational Engineering, Finance, and Science (cs.CE); Computational Geometry (cs.CG)

    Airborne LiDAR point cloud of a forest contains three dimensional data, from
    which vertical stand structure (including information about under-story trees)
    can be derived. This paper presents a segmentation approach for multi-story
    stands that strips the point cloud to its canopy layers, identifies individual
    tree segments within each layer using a DSM-based tree identification method as
    a building block, and combines the segments of immediate layers in order to fix
    potential over-segmentation of tree crowns across the layers. We introduce
    local layering that analyzes the vertical distributions of LiDAR points within
    their local neighborhoods in order to locally determine the height thresholds
    for layering the canopy. Unlike the previous work that stripped stiff layers
    within constrained areas, the local layering method strips flexible (in
    thickness and elevation) and narrower canopy layers within unconstrained areas.
    Statistical analyses showed that layering in general strongly improves
    identifying (specifically under-story) trees for the cost of moderately
    increasing over-segmentation rate of the identified trees. Combining tree
    segments across the immediate layers did not seem to improve tree
    identification accuracy remarkably, suggesting that layers separated canopy
    layers rather precisely.

    Improved Stereo Matching with Constant Highway Networks and Reflective Confidence Learning

    Amit Shaked, Lior Wolf
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    We present an improved three-step pipeline for the stereo matching problem
    and introduce multiple novelties at each stage. We propose a new highway
    network architecture for computing the matching cost at each possible
    disparity, based on multilevel weighted residual shortcuts, trained with a
    hybrid loss that supports multilevel comparison of image patches. A novel
    post-processing step is then introduced, which employs a second deep
    convolutional neural network for pooling global information from multiple
    disparities. This network outputs both the image disparity map, which replaces
    the conventional “winner takes all” strategy, and a confidence in the
    prediction. The confidence score is achieved by training the network with a new
    technique that we call the reflective loss. Lastly, the learned confidence is
    employed in order to better detect outliers in the refinement step. The
    proposed pipeline achieves state of the art accuracy on the largest and most
    competitive stereo benchmarks, and the learned confidence is shown to
    outperform all existing alternatives.

    EgoCap: Egocentric Marker-less Motion Capture with Two Fisheye Cameras (Extended Abstract)

    Helge Rhodin, Christian Richardt, Dan Casas, Eldar Insafutdinov, Mohammad Shafiei, Hans-Peter Seidel, Bernt Schiele, Christian Theobalt
    Comments: Short version of a SIGGRAPH Asia 2016 paper arXiv:1609.07306, presented at EPIC@ECCV16
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Marker-based and marker-less optical skeletal motion-capture methods use an
    outside-in arrangement of cameras placed around a scene, with viewpoints
    converging on the center. They often create discomfort by possibly needed
    marker suits, and their recording volume is severely restricted and often
    constrained to indoor scenes with controlled backgrounds. We therefore propose
    a new method for real-time, marker-less and egocentric motion capture which
    estimates the full-body skeleton pose from a lightweight stereo pair of fisheye
    cameras that are attached to a helmet or virtual-reality headset. It combines
    the strength of a new generative pose estimation framework for fisheye views
    with a ConvNet-based body-part detector trained on a new automatically
    annotated and augmented dataset. Our inside-in method captures full-body motion
    in general indoor and outdoor scenes, and also crowded scenes.

    p-DLA: A Predictive System Model for Onshore Oil and Gas Pipeline Dataset Classification and Monitoring – Part 1

    E.N. Osegi
    Comments: Working Paper
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    With the rise in militant activity and rogue behaviour in oil and gas regions
    around the world, oil pipeline disturbances is on the increase leading to huge
    losses to multinational operators and the countries where such facilities
    exist. However, this situation can be averted if adequate predictive monitoring
    schemes are put in place. We propose in the first part of this paper, an
    artificial intelligence predictive monitoring system capable of predictive
    classification and pattern recognition of pipeline datasets. The predictive
    system is based on a highly sparse predictive Deviant Learning Algorithm
    (p-DLA) designed to synthesize a sequence of memory predictive clusters for
    eventual monitoring, control and decision making. The DLA (p-DLA) is compared
    with a popular machine learning algorithm, the Long Short-Term Memory (LSTM)
    which is based on a temporal version of the standard feed-forward
    back-propagation trained artificial neural networks (ANNs). The results of
    simulations study show impressive results and validates the sparse memory
    predictive approach which favours the sub-synthesis of a highly compressed and
    low dimensional knowledge discovery and information prediction scheme. It also
    shows that the proposed new approach is competitive with a well-known and
    proven AI approach such as the LSTM.

    Super-Resolution Reconstruction of Electrical Impedance Tomography Images

    R. A. Borsoi, J. C. C. Aya, G. H. Costa, J. C. M. Bermudez
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Electrical Impedance Tomography (EIT) systems are becoming popular because
    they present several advantages over competing systems. However, EIT leads to
    images with very low resolution. Moreover, the nonuniform sampling
    characteristic of EIT precludes the straightforward application of traditional
    image ruper-resolution techniques. In this work, we propose a resampling based
    Super-Resolution method for EIT image quality improvement. Preliminary results
    show that the proposed technique can lead to substantial improvements in EIT
    image resolution, making it more competitive with other technologies.

    Deep-HiTS: Rotation Invariant Convolutional Neural Network for Transient Detection

    Guillermo Cabrera-Vives, Ignacio Reyes, Francisco Förster, Pablo A. Estévez, Juan-Carlos Maureira
    Journal-ref: The Astrophysical Journal, 2017
    Subjects: Instrumentation and Methods for Astrophysics (astro-ph.IM); Computer Vision and Pattern Recognition (cs.CV)

    We introduce Deep-HiTS, a rotation invariant convolutional neural network
    (CNN) model for classifying images of transients candidates into artifacts or
    real sources for the High cadence Transient Survey (HiTS). CNNs have the
    advantage of learning the features automatically from the data while achieving
    high performance. We compare our CNN model against a feature engineering
    approach using random forests (RF). We show that our CNN significantly
    outperforms the RF model reducing the error by almost half. Furthermore, for a
    fixed number of approximately 2,000 allowed false transient candidates per
    night we are able to reduce the miss-classified real transients by
    approximately 1/5. To the best of our knowledge, this is the first time CNNs
    have been used to detect astronomical transient events. Our approach will be
    very useful when processing images from next generation instruments such as the
    Large Synoptic Survey Telescope (LSST). We have made all our code and data
    available to the community for the sake of allowing further developments and
    comparisons at this https URL

    Assessing Uncertainties in X-ray Single-particle Three-dimensional reconstructions

    Stefan Engblom, Carl Nettelblad, Jing Liu
    Comments: 21 pages
    Subjects: Methodology (stat.ME); Computer Vision and Pattern Recognition (cs.CV); Data Analysis, Statistics and Probability (physics.data-an)

    Modern technology for producing extremely bright and coherent X-ray laser
    pulses provides the possibility to acquire a large number of diffraction
    patterns from individual biological nanoparticles, including proteins, viruses,
    and DNA. These two-dimensional diffraction patterns can be practically
    reconstructed and retrieved down to a resolution of a few angstrom. In
    principle, a sufficiently large collection of diffraction patterns will contain
    the required information for a full three-dimensional reconstruction of the
    biomolecule. The computational methodology for this reconstruction task is
    still under development and highly resolved reconstructions have not yet been
    produced.

    We analyze the Expansion-Maximization-Compression scheme, the current state
    of the art approach for this very challenging application, by isolating
    different sources of uncertainty. Through numerical experiments on synthetic
    data we evaluate their respective impact. We reach conclusions of relevance for
    handling actual experimental data, as well as pointing out certain improvements
    to the underlying estimation algorithm.

    We also introduce a practically applicable computational methodology in the
    form of bootstrap procedures for assessing reconstruction uncertainty in the
    real data case. We evaluate the sharpness of this approach and argue that this
    type of procedure will be critical in the near future when handling the
    increasing amount of data.


    Artificial Intelligence

    Conceptual Spaces for Cognitive Architectures: A Lingua Franca for Different Levels of Representation

    Antonio Lieto, Antonio Chella, Marcello Frixione
    Comments: 31 pages, 3 figures in Biologically Inspired Cognitive Architectures, 2017
    Subjects: Artificial Intelligence (cs.AI)

    During the last decades, many cognitive architectures (CAs) have been
    realized adopting different assumptions about the organization and the
    representation of their knowledge level. Some of them (e.g. SOAR [Laird
    (2012)]) adopt a classical symbolic approach, some (e.g. LEABRA [O’Reilly and
    Munakata (2000)]) are based on a purely connectionist model, while others (e.g.
    CLARION [Sun (2006)] adopt a hybrid approach combining connectionist and
    symbolic representational levels. Additionally, some attempts (e.g. biSOAR)
    trying to extend the representational capacities of CAs by integrating
    diagrammatical representations and reasoning are also available [Kurup and
    Chandrasekaran (2007)]. In this paper we propose a reflection on the role that
    Conceptual Spaces, a framework developed by Peter G”ardenfors [G”ardenfors
    (2000)] more than fifteen years ago, can play in the current development of the
    Knowledge Level in Cognitive Systems and Architectures. In particular, we claim
    that Conceptual Spaces offer a lingua franca that allows to unify and
    generalize many aspects of the symbolic, sub-symbolic and diagrammatic
    approaches (by overcoming some of their typical problems) and to integrate them
    on a common ground. In doing so we extend and detail some of the arguments
    explored by G”ardenfors [G”ardenfors (1997)] for defending the need of a
    conceptual, intermediate, representation level between the symbolic and the
    sub-symbolic one.

    An affective computational model for machine consciousness

    Rohitash Chandra
    Comments: under review
    Subjects: Artificial Intelligence (cs.AI)

    In the past, several models of consciousness have become popular and have led
    to the development of models for machine consciousness with varying degrees of
    success and challenges for simulation and implementations. Moreover, affective
    computing attributes that involve emotions, behavior and personality have not
    been the focus of models of consciousness as they lacked motivation for
    deployment in software applications and robots. The affective attributes are
    important factors for the future of machine consciousness with the rise of
    technologies that can assist humans. Personality and affection hence can give
    an additional flavor for the computational model of consciousness in humanoid
    robotics. Recent advances in areas of machine learning with a focus on deep
    learning can further help in developing aspects of machine consciousness in
    areas that can better replicate human sensory perceptions such as speech
    recognition and vision. With such advancements, one encounters further
    challenges in developing models that can synchronize different aspects of
    affective computing. In this paper, we review some existing models of
    consciousnesses and present an affective computational model that would enable
    the human touch and feel for robotic systems.

    STRIPS Planning in Infinite Domains

    Caelan Reed Garrett, Tomás Lozano-Pérez, Leslie Pack Kaelbling
    Comments: 10 pages
    Subjects: Artificial Intelligence (cs.AI)

    Many practical planning applications involve continuous quantities with
    non-linear constraints, which cannot be modeled using modern planners that
    construct a propositional representation. We introduce STRIPStream: an
    extension of the STRIPS language which supports infinite streams of objects and
    static predicates and provide two algorithms, which reduce the original problem
    to a sequence of finite-domain planning problems. The representation and
    algorithms are entirely domain independent. We demonstrate them on simple
    illustrative domains, and then on a high-dimensional, continuous robotic task
    and motion planning problem.

    Lazily Adapted Constant Kinky Inference for Nonparametric Regression and Model-Reference Adaptive Control

    Jan-Peter Calliess
    Subjects: Optimization and Control (math.OC); Artificial Intelligence (cs.AI); Learning (cs.LG); Systems and Control (cs.SY); Machine Learning (stat.ML)

    Techniques known as Nonlinear Set Membership prediction, Lipschitz
    Interpolation or Kinky Inference are approaches to machine learning that
    utilise presupposed Lipschitz properties to compute inferences over unobserved
    function values. Provided a bound on the true best Lipschitz constant of the
    target function is known a priori they offer convergence guarantees as well as
    bounds around the predictions. Considering a more general setting that builds
    on Hoelder continuity relative to pseudo-metrics, we propose an online method
    for estimating the Hoelder constant online from function value observations
    that possibly are corrupted by bounded observational errors. Utilising this to
    compute adaptive parameters within a kinky inference rule gives rise to a
    nonparametric machine learning method, for which we establish strong universal
    approximation guarantees. That is, we show that our prediction rule can learn
    any continuous function in the limit of increasingly dense data to within a
    worst-case error bound that depends on the level of observational uncertainty.
    We apply our method in the context of nonparametric model-reference adaptive
    control (MRAC). Across a range of simulated aircraft roll-dynamics and
    performance metrics our approach outperforms recently proposed alternatives
    that were based on Gaussian processes and RBF-neural networks. For
    discrete-time systems, we provide stability guarantees for our learning-based
    controllers both for the batch and the online learning setting.

    RNN-based Encoder-decoder Approach with Word Frequency Estimation

    Jun Suzuki, Masaaki Nagata
    Comments: 10 pages
    Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)

    This paper tackles the reduction of redundant repeating generation that is
    often observed in RNN-based encoder-decoder models. Our basic idea is to
    jointly estimate the upper-bound frequency of each target vocabulary in the
    encoder and control the output words based on the estimation in the decoder.
    Our method shows significant improvement over a strong RNN-based
    encoder-decoder baseline and achieved its best results on an abstractive
    summarization benchmark.

    Learning Weighted Association Rules in Human Phenotype Ontology

    Pietro Hiram Guzzi, Giuseppe Agapito, Marianna Milano, Mario Cannataro
    Subjects: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI)

    The Human Phenotype Ontology (HPO) is a structured repository of concepts
    (HPO Terms) that are associated to one or more diseases. The process of
    association is referred to as annotation. The relevance and the specificity of
    both HPO terms and annotations are evaluated by a measure defined as
    Information Content (IC). The analysis of annotated data is thus an important
    challenge for bioinformatics. There exist different approaches of analysis.
    From those, the use of Association Rules (AR) may provide useful knowledge, and
    it has been used in some applications, e.g. improving the quality of
    annotations. Nevertheless classical association rules algorithms do not take
    into account the source of annotation nor the importance yielding to the
    generation of candidate rules with low IC. This paper presents HPO-Miner (Human
    Phenotype Ontology-based Weighted Association Rules) a methodology for
    extracting Weighted Association Rules. HPO-Miner can extract relevant rules
    from a biological point of view. A case study on using of HPO-Miner on publicly
    available HPO annotation datasets is used to demonstrate the effectiveness of
    our methodology.

    Non-Negative Matrix Factorization Test Cases

    Connor Sell, Jeremy Kepner
    Comments: 4 pages, 3 figures, to appear in the proceedings of the 2015 IEEE MIT Undergraduate Research Conference
    Subjects: Numerical Analysis (math.NA); Artificial Intelligence (cs.AI); Numerical Analysis (cs.NA)

    Non-negative matrix factorization (NMF) is a prob- lem with many
    applications, ranging from facial recognition to document clustering. However,
    due to the variety of algorithms that solve NMF, the randomness involved in
    these algorithms, and the somewhat subjective nature of the problem, there is
    no clear “correct answer” to any particular NMF problem, and as a result, it
    can be hard to test new algorithms. This paper suggests some test cases for NMF
    algorithms derived from matrices with enumerable exact non-negative
    factorizations and perturbations of these matrices. Three algorithms using
    widely divergent approaches to NMF all give similar solutions over these test
    cases, suggesting that these test cases could be used as test cases for
    implementations of these existing NMF algorithms as well as potentially new NMF
    algorithms. This paper also describes how the proposed test cases could be used
    in practice.

    Digital Advertising Traffic Operation: Machine Learning for Process Discovery

    Massimiliano Dal Mas
    Comments: 6 pages; for details see: this this http URL
    Subjects: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)

    In a Web Advertising Traffic Operation it’s necessary to manage the
    day-to-day trafficking, pacing and optimization of digital and paid social
    campaigns. The data analyst on Traffic Operation can not only quickly provide
    answers but also speaks the language of the Process Manager and visually
    displays the discovered process problems. In order to solve a growing number of
    complaints in the customer service process, the weaknesses in the process
    itself must be identified and communicated to the department. With the help of
    Process Mining for the CRM data it is possible to identify unwanted loops and
    delays in the process. With this paper we propose a process discovery based on
    Machine Learning technique to automatically discover variations and detect at
    first glance what the problem is, and undertake corrective measures.


    Information Retrieval

    Patent Retrieval: A Literature Review

    Walid Shalaby, Wlodek Zadrozny
    Subjects: Information Retrieval (cs.IR)

    With the ever increasing number of filed patent applications every year, the
    need for effective and efficient systems for managing such tremendous amounts
    of data becomes inevitably important. Patent Retrieval (PR) is considered is
    the pillar of almost all patent analysis tasks. PR is a subfield of Information
    Retrieval (IR) which is concerned with developing techniques and methods that
    effectively and efficiently retrieve relevant patent documents in response to a
    given search request. In this paper we present a comprehensive review on PR
    methods and approaches. It is clear that, recent successes and maturity in IR
    applications such as Web search can not be transferred directly to PR without
    deliberate domain adaptation and customization. Furthermore, state-of-the-art
    performance in automatic PR is still around average. These observations
    motivates the need for interactive search tools which provide cognitive
    assistance to patent professionals with minimal effort. These tools must also
    be developed in hand with patent professionals considering their practices and
    expectations. We additionally touch on related tasks to PR such as patent
    valuation, litigation, licensing, and highlight potential opportunities and
    open directions for computational scientists in these domains.

    Interactive Movie Recommendation Through Latent Semantic Analysis and Storytelling

    Kodzo Wegba, Aidong Lu, Yuemeng Li, Wencheng Wang
    Comments: 10 pages
    Subjects: Information Retrieval (cs.IR); Social and Information Networks (cs.SI)

    Recommendation has become one of the most important components of online
    services for improving sale records, however visualization work for online
    recommendation is still very limited. This paper presents an interactive
    recommendation approach with the following two components. First, rating
    records are the most widely used data for online recommendation, but they are
    often processed in high-dimensional spaces that can not be easily understood or
    interacted with. We propose a Latent Semantic Model (LSM) that captures the
    statistical features of semantic concepts on 2D domains and abstracts user
    preferences for personal recommendation. Second, we propose an interactive
    recommendation approach through a storytelling mechanism for promoting the
    communication between the user and the recommendation system. Our approach
    emphasizes interactivity, explicit user input, and semantic information convey;
    thus it can be used by general users without any knowledge of recommendation or
    visualization algorithms. We validate our model with data statistics and
    demonstrate our approach with case studies from the MovieLens100K dataset. Our
    approaches of latent semantic analysis and interactive recommendation can also
    be extended to other network-based visualization applications, including
    various online recommendation systems.

    Self-Taught Convolutional Neural Networks for Short Text Clustering

    Jiaming Xu, Bo Xu, Peng Wang, Suncong Zheng, Guanhua Tian, Jun Zhao, Bo Xu
    Comments: 33 pages, accepted for publication in Neural Networks
    Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL)

    Short text clustering is a challenging problem due to its sparseness of text
    representation. Here we propose a flexible Self-Taught Convolutional neural
    network framework for Short Text Clustering (dubbed STC^2), which can flexibly
    and successfully incorporate more useful semantic features and learn non-biased
    deep text representation in an unsupervised manner. In our framework, the
    original raw text features are firstly embedded into compact binary codes by
    using one existing unsupervised dimensionality reduction methods. Then, word
    embeddings are explored and fed into convolutional neural networks to learn
    deep feature representations, meanwhile the output units are used to fit the
    pre-trained binary codes in the training process. Finally, we get the optimal
    clusters by employing K-means to cluster the learned representations. Extensive
    experimental results demonstrate that the proposed framework is effective,
    flexible and outperform several popular clustering methods when tested on three
    public short text datasets.

    Integrating sentiment and social structure to determine preference alignments: The Irish Marriage Referendum

    David J.P. O'Sullivan, Guillermo Garduño-Hernández, James P. Gleeson, Mariano Beguerisse-Díaz
    Comments: 16 pages, 12 figures
    Subjects: Social and Information Networks (cs.SI); Computation and Language (cs.CL); Information Retrieval (cs.IR); Physics and Society (physics.soc-ph)

    We investigate the relationship between social structure and sentiment
    through the analysis of half a million tweets about the Irish Marriage
    Referendum of 2015. We obtain the sentiment of every tweet with the hashtags
    #marref and #marriageref posted in the days leading to the referendum, and
    construct networks to aggregate sentiment and study the interactions among
    users. The sentiment of the mention tweets that a user sends is correlated with
    the sentiment of the mentions received, and there are significantly more
    connections between users with similar sentiment scores than among users with
    opposite scores. We combine the community structure of the follower and mention
    networks, the activity level of the users, and sentiment scores to find groups
    of users who support voting ‘yes’ or ‘no’ on the referendum. We find that many
    conversations between users on opposing sides of the debate occurred in the
    absence of follower connections, suggesting that there were efforts by some
    users to establish dialogue and debate across ideological divisions. These
    results show that social structures can be successfully integrated with
    sentiment to analyse and understand the disposition of social media users. We
    discuss the implications of our work for the integration of data and meta-data,
    opinion dynamics, public opinion modelling and polling.


    Computation and Language

    Aspect-augmented Adversarial Networks for Domain Adaptation

    Yuan Zhang, Regina Barzilay, Tommi Jaakkola
    Comments: TACL
    Subjects: Computation and Language (cs.CL)

    We introduce a neural method for transfer learning between two (source and
    target) classification tasks or aspects over the same domain. Instead of target
    labels, we assume a few keywords pertaining to source and target aspects
    indicating sentence relevance rather than document class labels. Documents are
    encoded by learning to embed and softly select relevant sentences in an
    aspect-dependent manner. A shared classifier is trained on the source encoded
    documents and labels, and applied to target encoded documents. We ensure
    transfer through aspect-adversarial training so that encoded documents are, as
    sets, aspect-invariant. Experimental results demonstrate that our approach
    outperforms different baselines and model variants on two datasets, yielding an
    improvement of 24% on a pathology dataset and 5% on a review dataset.

    Social Media Argumentation Mining: The Quest for Deliberateness in Raucousness

    Jan Šnajder
    Subjects: Computation and Language (cs.CL)

    Argumentation mining from social media content has attracted increasing
    attention. The task is both challenging and rewarding. The informal nature of
    user-generated content makes the task dauntingly difficult. On the other hand,
    the insights that could be gained by a large-scale analysis of social media
    argumentation make it a very worthwhile task. In this position paper I discuss
    the motivation for social media argumentation mining, as well as the tasks and
    challenges involved.

    Expanding Subjective Lexicons for Social Media Mining with Embedding Subspaces

    Silvio Amir, Rámon Astudillo, Wang Ling, Paula C. Carvalho, Mário J. Silva
    Subjects: Computation and Language (cs.CL)

    Recent approaches for sentiment lexicon induction have capitalized on
    pre-trained word embeddings that capture latent semantic properties. However,
    embeddings obtained by optimizing performance of a given task (e.g. predicting
    contextual words) are sub-optimal for other applications. In this paper, we
    address this problem by exploiting task-specific representations, induced via
    embedding sub-space projection. This allows us to expand lexicons describing
    multiple semantic properties. For each property, our model jointly learns
    suitable representations and the concomitant predictor. Experiments conducted
    over multiple subjective lexicons, show that our model outperforms previous
    work and other baselines; even in low training data regimes. Furthermore,
    lexicon-based sentiment classifiers built on top of our lexicons outperform
    similar resources and yield performances comparable to those of supervised
    models.

    RNN-based Encoder-decoder Approach with Word Frequency Estimation

    Jun Suzuki, Masaaki Nagata
    Comments: 10 pages
    Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)

    This paper tackles the reduction of redundant repeating generation that is
    often observed in RNN-based encoder-decoder models. Our basic idea is to
    jointly estimate the upper-bound frequency of each target vocabulary in the
    encoder and control the output words based on the estimation in the decoder.
    Our method shows significant improvement over a strong RNN-based
    encoder-decoder baseline and achieved its best results on an abstractive
    summarization benchmark.

    A POS Tagger for Code Mixed Indian Social Media Text – ICON-2016 NLP Tools Contest Entry from Surukam

    Sree Harsha Ramesh, Raveena R Kumar
    Comments: 4 Pages, 13th International Conference on Natural Language Processing, Varanasi, India
    Subjects: Computation and Language (cs.CL)

    Building Part-of-Speech (POS) taggers for code-mixed Indian languages is a
    particularly challenging problem in computational linguistics due to a dearth
    of accurately annotated training corpora. ICON, as part of its NLP tools
    contest has organized this challenge as a shared task for the second
    consecutive year to improve the state-of-the-art. This paper describes the POS
    tagger built at Surukam to predict the coarse-grained and fine-grained POS tags
    for three language pairs – Bengali-English, Telugu-English and Hindi-English,
    with the text spanning three popular social media platforms – Facebook,
    WhatsApp and Twitter. We employed Conditional Random Fields as the sequence
    tagging algorithm and used a library called sklearn-crfsuite – a thin wrapper
    around CRFsuite for training our model. Among the features we used include –
    character n-grams, language information and patterns for emoji, number,
    punctuation and web-address. Our submissions in the constrained
    environment,i.e., without making any use of monolingual POS taggers or the
    like, obtained an overall average F1-score of 76.45%, which is comparable to
    the 2015 winning score of 76.79%.

    Integrating sentiment and social structure to determine preference alignments: The Irish Marriage Referendum

    David J.P. O'Sullivan, Guillermo Garduño-Hernández, James P. Gleeson, Mariano Beguerisse-Díaz
    Comments: 16 pages, 12 figures
    Subjects: Social and Information Networks (cs.SI); Computation and Language (cs.CL); Information Retrieval (cs.IR); Physics and Society (physics.soc-ph)

    We investigate the relationship between social structure and sentiment
    through the analysis of half a million tweets about the Irish Marriage
    Referendum of 2015. We obtain the sentiment of every tweet with the hashtags
    #marref and #marriageref posted in the days leading to the referendum, and
    construct networks to aggregate sentiment and study the interactions among
    users. The sentiment of the mention tweets that a user sends is correlated with
    the sentiment of the mentions received, and there are significantly more
    connections between users with similar sentiment scores than among users with
    opposite scores. We combine the community structure of the follower and mention
    networks, the activity level of the users, and sentiment scores to find groups
    of users who support voting ‘yes’ or ‘no’ on the referendum. We find that many
    conversations between users on opposing sides of the debate occurred in the
    absence of follower connections, suggesting that there were efforts by some
    users to establish dialogue and debate across ideological divisions. These
    results show that social structures can be successfully integrated with
    sentiment to analyse and understand the disposition of social media users. We
    discuss the implications of our work for the integration of data and meta-data,
    opinion dynamics, public opinion modelling and polling.

    Self-Taught Convolutional Neural Networks for Short Text Clustering

    Jiaming Xu, Bo Xu, Peng Wang, Suncong Zheng, Guanhua Tian, Jun Zhao, Bo Xu
    Comments: 33 pages, accepted for publication in Neural Networks
    Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL)

    Short text clustering is a challenging problem due to its sparseness of text
    representation. Here we propose a flexible Self-Taught Convolutional neural
    network framework for Short Text Clustering (dubbed STC^2), which can flexibly
    and successfully incorporate more useful semantic features and learn non-biased
    deep text representation in an unsupervised manner. In our framework, the
    original raw text features are firstly embedded into compact binary codes by
    using one existing unsupervised dimensionality reduction methods. Then, word
    embeddings are explored and fed into convolutional neural networks to learn
    deep feature representations, meanwhile the output units are used to fit the
    pre-trained binary codes in the training process. Finally, we get the optimal
    clusters by employing K-means to cluster the learned representations. Extensive
    experimental results demonstrate that the proposed framework is effective,
    flexible and outperform several popular clustering methods when tested on three
    public short text datasets.


    Distributed, Parallel, and Cluster Computing

    A Resource Management Protocol for Mobile Cloud Using Auto-Scaling

    Chathura Sarathchandra Magurawalage, Kun Yang, Ritoa Patrik, Michael Georgiades, Kezhi Wang
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    Cloud radio access network (C-RAN) and Mobile Cloud Computing (MCC) have
    emerged as promising candidates for the next generation access network
    techniques. MCC offers resource limited mobile devices to offload
    computationally intensive tasks to the cloud, while C-RAN offers a technology
    that addresses increasing mobile traffic. In this paper, we propose a protocol
    that allows task offloading and managing resources in both C-RAN and mobile
    cloud together using a centralised controller. Experiments on resource
    management using cloud auto-scaling shows that resource (CPU, RAM, Storage)
    scaling times vary.

    Multi-objective dynamic virtual machine consolidation in the cloud using ant colony system

    Adnan Ashraf, Ivan Porres
    Comments: The manuscript has been accepted for publication in the International Journal of Parallel, Emergent and Distributed Systems
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    In this paper, we present a novel multi-objective ant colony system algorithm
    for virtual machine (VM) consolidation in cloud data centers. The proposed
    algorithm builds VM migration plans, which are then used to minimize
    over-provisioning of physical machines (PMs) by consolidating VMs on
    under-utilized PMs. It optimizes two objectives that are ordered by their
    importance. The first and foremost objective in the proposed algorithm is to
    maximize the number of released PMs. Moreover, since VM migration is a
    resource-intensive operation, it also tries to minimize the number of VM
    migrations. The proposed algorithm is empirically evaluated in a series of
    experiments. The experimental results show that the proposed algorithm provides
    an efficient solution for VM consolidation in cloud data centers. Moreover, it
    outperforms two existing ant colony optimization based VM consolidation
    algorithms in terms of number of released PMs and number of VM migrations.

    Analysis of a Stochastic Model of Replication in Large Distributed Storage Systems: A Mean-Field Approach

    Wen Sun, Véronique Simon, Sébastien Monnet, Philippe Robert, Pierre Sens
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    Distributed storage systems such as Hadoop File System or Google File System
    (GFS) ensure data availability and durability using replication. This paper is
    focused on the analysis of the efficiency of replication mechanism that
    determines the location of the copies of a given file at some server. The
    variability of the loads of the nodes of the network is investigated for
    several policies. Three replication mechanisms are tested against simulations
    in the context of a real implementation of a such a system: Random, Least
    Loaded and Power of Choice.

    The simulations show that some of these policies may lead to quite unbalanced
    situations: if )eta( is the average number of copies per node it turns out
    that, at equilibrium, the load of the nodes may exhibit a high variability. It
    is shown in this paper that a simple variant of a power of choice type
    algorithm has a striking effect on the loads of the nodes: at equilibrium, the
    load of a node has a bounded support, almost all nodes have a load less than
    )2eta(.

    Mathematical models are introduced and investigated to explain this unusual,
    quite surprising, phenomenon. Our study relies on probabilistic methods,
    mean-field analysis, to analyze the asymptotic behavior of an arbitrary node of
    the network when the total number of nodes gets large. An additional ingredient
    is the use of stochastic calculus with marked Poisson point processes to
    establish some of our results.

    Packet Latency of Deterministic Broadcasting in Adversarial Multiple Access Channels

    Lakshmi Anantharamu, Bogdan S. Chlebus, Dariusz R. Kowalski, Mariusz A. Rokicki
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    We study broadcasting on multiple access channels with dynamic packet
    arrivals and jamming. The communication environments is represented by
    adversarial models which specify constraints on packet arrivals and jamming. We
    consider deterministic distributed broadcast algorithms and give upper bounds
    on the worst-case packet latency and the number of queued packets in relation
    to the parameters defining adversaries. Packet arrivals are determined by the
    rate of injections and number of packets that can arrive in one round. Jamming
    is constrained by the rate with which the adversary can jam rounds and by the
    number of consecutive rounds that can be jammed.

    A scalable approach for tree segmentation within small-footprint airborne LiDAR data

    Hamid Hamraz, Marco A. Contreras, Jun Zhang
    Comments: Highlights: – A scalable distributed approach for tree segmentation was developed and analyzed. – ~2 million trees in a 7440 ha forest was segmented in 2.5 hours using 192 cores. – 2% false positive trees were identified as a result of the distributed run. – The approach can be used to scale up processing other big spatial data
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Computational Engineering, Finance, and Science (cs.CE)

    This paper presents a distributed approach that scales up to segment tree
    crowns within a LiDAR point cloud representing an arbitrarily large forested
    area. The approach uses a single-processor tree segmentation algorithm as a
    building block in order to process the data delivered in the shape of tiles in
    parallel. The distributed processing is performed in a master-slave manner, in
    which the master maintains the global map of the tiles and coordinates the
    slaves that segment tree crowns within and across the boundaries of the tiles.
    A minimal bias was introduced to the number of detected trees because of trees
    lying across the tile boundaries, which was quantified and adjusted for.
    Theoretical and experimental analyses of the runtime of the approach revealed a
    near linear speedup. The estimated number of trees categorized by crown class
    and the associated error margins as well as the height distribution of the
    detected trees aligned well with field estimations, verifying that the
    distributed approach works correctly. The approach enables providing
    information of individual tree locations and point cloud segments for a
    forest-level area in a timely manner, which can be used to create detailed
    remotely sensed forest inventories. Although the approach was presented for
    tree segmentation within LiDAR point clouds, the idea can also be generalized
    to scale up processing other big spatial datasets.


    Learning

    Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Selective Execution

    Lanlan Liu, Jia Deng
    Comments: CVPR 2017 Submission
    Subjects: Learning (cs.LG); Machine Learning (stat.ML)

    We introduce Dynamic Deep Neural Networks (D2NN), a new type of feed-forward
    deep neural network that allow selective execution. Given an input, only a
    subset of D2NN neurons are executed, and the particular subset is determined by
    the D2NN itself. By pruning unnecessary computation depending on input, D2NNs
    provide a way to improve computational efficiency. To achieve dynamic selective
    execution, a D2NN augments a regular feed-forward deep neural network (directed
    acyclic graph of differentiable modules) with one or more controller modules.
    Each controller module is a sub-network whose output is a decision that
    controls whether other modules can execute. A D2NN is trained end to end. Both
    regular modules and controller modules in a D2NN are learnable and are jointly
    trained to optimize both accuracy and efficiency. Such training is achieved by
    integrating backpropagation with reinforcement learning. With extensive
    experiments of various D2NN architectures on image classification tasks, we
    demonstrate that D2NNs are general and flexible, and can effectively optimize
    accuracy-efficiency trade-offs.

    Outlier Robust Online Learning

    Jiashi Feng, Huan Xu, Shie Mannor
    Subjects: Learning (cs.LG); Machine Learning (stat.ML)

    We consider the problem of learning from noisy data in practical settings
    where the size of data is too large to store on a single machine. More
    challenging, the data coming from the wild may contain malicious outliers. To
    address the scalability and robustness issues, we present an online robust
    learning (ORL) approach. ORL is simple to implement and has provable robustness
    guarantee — in stark contrast to existing online learning approaches that are
    generally fragile to outliers. We specialize the ORL approach for two concrete
    cases: online robust principal component analysis and online linear regression.
    We demonstrate the efficiency and robustness advantages of ORL through
    comprehensive simulations and predicting image tags on a large-scale data set.
    We also discuss extension of the ORL to distributed learning and provide
    experimental evaluations.

    Classification of Smartphone Users Using Internet Traffic

    Andrey Finkelstein, Ron Biton, Rami Puzis, Asaf Shabtai
    Subjects: Learning (cs.LG); Cryptography and Security (cs.CR)

    Today, smartphone devices are owned by a large portion of the population and
    have become a very popular platform for accessing the Internet. Smartphones
    provide the user with immediate access to information and services. However,
    they can easily expose the user to many privacy risks. Applications that are
    installed on the device and entities with access to the device’s Internet
    traffic can reveal private information about the smartphone user and steal
    sensitive content stored on the device or transmitted by the device over the
    Internet. In this paper, we present a method to reveal various demographics and
    technical computer skills of smartphone users by their Internet traffic
    records, using machine learning classification models. We implement and
    evaluate the method on real life data of smartphone users and show that
    smartphone users can be classified by their gender, smoking habits, software
    programming experience, and other characteristics.

    NIPS 2016 Tutorial: Generative Adversarial Networks

    Ian Goodfellow
    Subjects: Learning (cs.LG)

    This report summarizes the tutorial presented by the author at NIPS 2016 on
    generative adversarial networks (GANs). The tutorial describes: (1) Why
    generative modeling is a topic worth studying, (2) how generative models work,
    and how GANs compare to other generative models, (3) the details of how GANs
    work, (4) research frontiers in GANs, and (5) state-of-the-art image models
    that combine GANs with other methods. Finally, the tutorial contains three
    exercises for readers to complete, and the solutions to these exercises.

    Lazily Adapted Constant Kinky Inference for Nonparametric Regression and Model-Reference Adaptive Control

    Jan-Peter Calliess
    Subjects: Optimization and Control (math.OC); Artificial Intelligence (cs.AI); Learning (cs.LG); Systems and Control (cs.SY); Machine Learning (stat.ML)

    Techniques known as Nonlinear Set Membership prediction, Lipschitz
    Interpolation or Kinky Inference are approaches to machine learning that
    utilise presupposed Lipschitz properties to compute inferences over unobserved
    function values. Provided a bound on the true best Lipschitz constant of the
    target function is known a priori they offer convergence guarantees as well as
    bounds around the predictions. Considering a more general setting that builds
    on Hoelder continuity relative to pseudo-metrics, we propose an online method
    for estimating the Hoelder constant online from function value observations
    that possibly are corrupted by bounded observational errors. Utilising this to
    compute adaptive parameters within a kinky inference rule gives rise to a
    nonparametric machine learning method, for which we establish strong universal
    approximation guarantees. That is, we show that our prediction rule can learn
    any continuous function in the limit of increasingly dense data to within a
    worst-case error bound that depends on the level of observational uncertainty.
    We apply our method in the context of nonparametric model-reference adaptive
    control (MRAC). Across a range of simulated aircraft roll-dynamics and
    performance metrics our approach outperforms recently proposed alternatives
    that were based on Gaussian processes and RBF-neural networks. For
    discrete-time systems, we provide stability guarantees for our learning-based
    controllers both for the batch and the online learning setting.

    Very Fast Kernel SVM under Budget Constraints

    David Picard
    Subjects: Machine Learning (stat.ML); Learning (cs.LG)

    In this paper we propose a fast online Kernel SVM algorithm under tight
    budget constraints. We propose to split the input space using LVQ and train a
    Kernel SVM in each cluster. To allow for online training, we propose to limit
    the size of the support vector set of each cluster using different strategies.
    We show in the experiment that our algorithm is able to achieve high accuracy
    while having a very high number of samples processed per second both in
    training and in the evaluation.

    Deep Neural Networks to Enable Real-time Multimessenger Astrophysics

    Daniel George, E. A. Huerta
    Comments: 20 pages, 15 figures
    Subjects: Instrumentation and Methods for Astrophysics (astro-ph.IM); Astrophysics of Galaxies (astro-ph.GA); High Energy Astrophysical Phenomena (astro-ph.HE); Learning (cs.LG); General Relativity and Quantum Cosmology (gr-qc)

    We introduce a new methodology for time-domain signal processing, based on
    deep learning neural networks, which has the potential to revolutionize data
    analysis in science. To illustrate how this enables real-time multimessenger
    astrophysics, we designed two deep convolutional neural networks that can
    analyze time-series data from observatories including advanced LIGO. The first
    neural network recognizes the presence of gravitational waves from binary black
    hole mergers, and the second one estimates the mass of each black hole, given
    weak signals hidden in extremely noisy time-series inputs. We highlight the
    advantages offered by this novel method, which outperforms matched-filtering or
    conventional machine learning techniques, and propose strategies to extend our
    implementation for simultaneously targeting different classes of gravitational
    wave sources while ignoring anomalous noise transients. Our results strongly
    indicate that deep neural networks are highly efficient and versatile tools for
    directly processing raw noisy data streams. Furthermore, we pioneer a new
    paradigm to accelerate scientific discovery by combining high-performance
    simulations on traditional supercomputers and artificial intelligence
    algorithms that exploit innovative hardware architectures such as
    deep-learning-optimized GPUs. This unique approach immediately provides a
    natural framework to unify multi-spectrum observations in real-time, thus
    enabling coincident detection campaigns of gravitational waves sources and
    their electromagnetic counterparts.


    Information Theory

    SINR Outage Evaluation in Cellular Networks: Saddle Point Approximation (SPA) Using Normal Inverse Gaussian (NIG) Distribution

    Sudarshan Guruacharya, Hina Tabassum, Ekram Hossain
    Comments: arXiv admin note: substantial text overlap with arXiv:1607.06887
    Subjects: Information Theory (cs.IT)

    Signal-to-noise-plus-interference ratio (SINR) outage probability is among
    one of the key performance metrics of a wireless cellular network. In this
    paper, we propose a semi-analytical method based on saddle point approximation
    (SPA) technique to calculate the SINR outage of a wireless system whose SINR
    can be modeled in the form )frac{sum_{i=1}^M X_i}{sum_{i=1}^N Y_i +1}( where
    )X_i( denotes the useful signal power, )Y_i( denotes the power of the
    interference signal, and )sum_{i=1}^M X_i(, )sum_{i=1}^N Y_i( are independent
    random variables. Both )M( and )N( can also be random variables. The proposed
    approach is based on the saddle point approximation to cumulative distribution
    function (CDF) as given by it{Wood-Booth-Butler formula}. The approach is
    applicable whenever the cumulant generating function (CGF) of the received
    signal and interference exists, and it allows us to tackle distributions with
    large skewness and kurtosis with higher accuracy. In this regard, we exploit a
    four parameter it{normal-inverse Gaussian} (NIG) distribution as a base
    distribution. Given that the skewness and kurtosis satisfy a specific
    condition, NIG-based SPA works reliably. When this condition is violated, we
    recommend SPA based on normal or symmetric NIG distribution, both special cases
    of NIG distribution, at the expense of reduced accuracy. For the purpose of
    demonstration, we apply SPA for the SINR outage evaluation of a typical user
    experiencing a downlink coordinated multi-point transmission (CoMP) from the
    base stations (BSs) that are modeled by homogeneous Poisson point process. We
    characterize the outage of the typical user in scenarios such as (a)~when the
    number and locations of interferers are random, and (b)~when the fading
    channels and number of interferers are random. Numerical results are presented
    to illustrate the accuracy of the proposed set of approximations.

    A time-variant channel prediction and feedback framework for interference alignment

    Zhinan Xu, Markus Hofer, Thomas Zemen
    Subjects: Information Theory (cs.IT)

    In interference channels, channel state information (CSI) can be exploited to
    reduce the interference signal dimensions and thus achieve the optimal capacity
    scaling, i.e. degrees of freedom, promised by the interference alignment
    technique. However, imperfect CSI, due to channel estimation error, imperfect
    CSI feedback and time selectivity of the channel, lead to a performance loss.
    In this work, we propose a novel limited feedback algorithm for single-input
    single-output interference alignment in time-variant channels. The feedback
    algorithm encodes the channel evolution in a small number of subspace
    coefficients, which allow for reduced-rank channel prediction to compensate for
    the channel estimation error due to time selectivity of the fading process and
    feedback delay. An upper bound for the rate loss caused by feedback
    quantization and channel prediction is derived. Based on this bound, we develop
    a dimension switching algorithm for the reduced-rank predictor to find the best
    tradeoff between quantization- and prediction-error. Besides, we characterize
    the scaling of the required number of feedback bits in order to decouple the
    rate loss due to channel quantization from the transmit power. Simulation
    results show that a rate gain over the traditional non-predictive feedback
    strategy can be secured and a 60% higher rate is achieved at 20 dB
    signal-to-noise ratio with moderate mobility.

    Beam-On-Graph: Simultaneous Channel Estimation in Multi-user Millimeter Wave MIMO Systems

    Matthew Kokshoorn, He Chen, Yonghui Li, Branka Vucetic
    Comments: Submitted for journal publication. arXiv admin note: substantial text overlap with arXiv:1612.02113
    Subjects: Information Theory (cs.IT)

    This paper is concerned with the channel estimation problem in multi-user
    millimeter wave (mmWave) wireless systems with large antenna arrays. We develop
    a novel simultaneous-estimation with iterative fountain training (SWIFT)
    framework, in which multiple users estimate their channels at the same time and
    the required number of channel measurements is adapted to various channel
    conditions of different users. To achieve this, we represent the beam direction
    estimation process by a graph, referred to as the beam-on-graph, and associate
    the channel estimation process with a code-on-graph decoding problem.
    Specifically, the base station (BS) and each user measure the channel with a
    series of random combinations of transmit/receive beamforming vectors until the
    channel estimate converges. As the proposed SWIFT does not adapt the BS’s beams
    to any single user, we are able to estimate all user channels simultaneously.
    Simulation results show that SWIFT can significantly outperform the existing
    random beamforming-based approaches, which use a predetermined number of
    measurements, over a wide range of signal-to-noise ratios and channel coherence
    time. Furthermore, by utilizing the users’ order in terms of completing their
    channel estimation, our SWIFT framework can infer the sequence of users’
    channel quality and perform effective user scheduling to achieve superior
    performance.

    Stochastic Geometry-based Comparison of Secrecy Enhancement Techniques in D2D Networks

    Mustafa A. Kishk, Harpreet S. Dhillon
    Subjects: Information Theory (cs.IT)

    This letter presents a performance comparison of two popular secrecy
    enhancement techniques in wireless networks: (i) creating guard zones by
    restricting transmissions of legitimate transmitters whenever any eavesdropper
    is detected in their vicinity, and (ii) adding artificial noise to the
    confidential messages to make it difficult for the eavesdroppers to decode
    them. Using tools from stochastic geometry, we first derive the secrecy outage
    probability at the eavesdroppers as well as the coverage probability at the
    legitimate users for both these techniques. Using these results, we derive a
    threshold on the density of the eavesdroppers below which no secrecy enhancing
    technique is required to ensure a target secrecy outage probability. For
    eavesdropper densities above this threshold, we concretely characterize the
    regimes in which each technique outperforms the other. Our results demonstrate
    that guard zone technique is better when the distances between the transmitters
    and their legitimate receivers are higher than a certain threshold.

    Self-Interference in Full-Duplex Multi-User MIMO Channels

    Arman Shojaeifard, Kai-Kit Wong, Marco Di Renzo, Gan Zheng, Khairi Ashour Hamdi, Jie Tang
    Subjects: Information Theory (cs.IT)

    We consider a multi-user multiple-input multiple-output (MIMO) setup where
    full-duplex (FD) multi-antenna nodes apply linear beamformers to simultaneously
    transmit and receive multiple streams over Rician fading channels. The exact
    first and second positive moments of the residual self-interference (SI),
    involving the squared norm of a sum of non-identically distributed random
    variables, are derived in closed-form. The method of moments is hence invoked
    to provide a Gamma approximation for the residual SI distribution. The proposed
    theorem holds under arbitrary linear precoder/decoder design, number of
    antennas and streams, and SI cancellation capability.

    Some Repeated-Root Constacyclic Codes over Galois Rings

    Hongwei Liu, Maouche Youcef
    Subjects: Information Theory (cs.IT)

    Codes over the Galois rings have been studied by many researchers, negacyclic
    codes over )GR(2^a,m)( of length )2^s( have been characterized by the fact that
    the ring ){cal R}_2(a,m,-1)= frac{GR(2^a,m)[x]}{langle x^{2^s}+1
    angle}( is
    a chain ring, furthermore, these results have been generalized to
    )lambda(-constacyclic codes for any unit )lambda( of the form )4z-1(, )zin
    GR(2^a, m)(. In this paper, we give more general cases and investigate all
    cases where ){cal R}_2(a,m,gamma)= frac{GR(2^a,m)[x]}{langle x^{2^s}-gamma

    angle}( is a chain ring, moreover, we give necessary and sufficient
    conditions for the ring ){cal R}_2(a,m,gamma)( to be a chain ring.
    Furthermore, we generalize these results to all odd prime number, by giving
    necessary and sufficient conditions for the ring ){cal
    R}_p(a,m,gamma)=frac{GR(p^a,m)[x]}{langle x^{p^s}-gamma
    angle}( to be a
    chain ring, using this structure we investigate all )gamma(-constacyclic codes
    over )GR(p^a,m)(, where ){cal R}_p(a,m,gamma)( is a chain ring. The dual
    codes and necessary and sufficient conditions for the existence of
    self-orthogonal and self-dual )gamma(-constacyclic codes are provided. Among
    others, for any prime )p(, the structure of ){cal
    R}_p(a,m,gamma)=frac{GR(p^a,m)[x]}{langle x^{p^s}-gamma
    angle}( is used to
    establish the Hamming and homogeneous distance.

    Access Strategy in Super WiFi Network Powered by Solar Energy Harvesting: A POMDP Method

    Tingwu Wang, Jian Wang, Chunxiao Jiang, Jingjing Wang, Yong Ren
    Subjects: Information Theory (cs.IT)

    The recently announced Super Wi-Fi Network proposal in United States is
    aiming to enable Internet access in a nation-wide area. As traditional
    cable-connected power supply system becomes impractical or costly for a wide
    range wireless network, new infrastructure deployment for Super Wi-Fi is
    required. The fast developing Energy Harvesting (EH) techniques receive global
    attentions for their potential of solving the above power supply problem. It is
    a critical issue, from the user’s perspective, how to make efficient network
    selection and access strategies. Unlike traditional wireless networks, the
    battery charge state and tendency in EH based networks have to be taken into
    account when making network selection and access, which has not been well
    investigated. In this paper, we propose a practical and efficient framework for
    multiple base stations access strategy in an EH powered Super Wi-Fi network. We
    consider the access strategy from the user’s perspective, who exploits downlink
    transmission opportunities from one base station. To formulate the problem, we
    used Partially Observable Markov Decision Process (POMDP) to model users’
    observations on the base stations’ battery situation and decisions on the base
    station selection and access. Simulation results show that our methods are
    efficacious and significantly outperform the traditional widely used CSMA
    method.

    Construction and Encoding of QC-LDPC Codes Using Group Rings

    Hassan Khodaiemehr, Dariush Kiani
    Comments: 56 pages, 9 figures. arXiv admin note: text overlap with arXiv:cs/0611112 by other authors
    Subjects: Information Theory (cs.IT)

    Quasi-cyclic (QC) low-density parity-check (LDPC) codes which are known as
    QC-LDPC codes, have many applications due to their simple encoding
    implementation by means of cyclic shift registers. In this paper, we construct
    QC-LDPC codes from group rings. A group ring is a free module (at the same time
    a ring) constructed in a natural way from any given ring and any given group.
    We present a structure based on the elements of a group ring for constructing
    QC-LDPC codes. Some of the previously addressed methods for constructing
    QC-LDPC codes based on finite fields are special cases of the proposed
    construction method. The constructed QC-LDPC codes perform very well over the
    additive white Gaussian noise (AWGN) channel with iterative decoding in terms
    of bit-error probability and block-error probability. Simulation results
    demonstrate that the proposed codes have competitive performance in comparison
    with the similar existing LDPC codes. Finally, we propose a new encoding method
    for the proposed group ring based QC-LDPC codes that can be implemented faster
    than the current encoding methods. The encoding complexity of the proposed
    method is analyzed mathematically, and indicates a significate reduction in the
    required number of operations, even when compared to the available efficient
    encoding methods that have linear time and space complexities.

    Interference Minimization in 5G Heterogeneous Networks

    Tao Han, Guoqiang Mao, Qiang Li, Lijun Wang, Jing Zhang
    Comments: 7 pages, 3 figures
    Journal-ref: Mobile Networks and Applications, vol. 20, no. 6, pp. 756-762,
    2015
    Subjects: Information Theory (cs.IT); Networking and Internet Architecture (cs.NI)

    In this paper, we focus on one of the representative 5G network scenarios,
    namely multi-tier heterogeneous cellular networks. User association is
    investigated in order to reduce the down-link co-channel interference. Firstly,
    in order to analyze the multi-tier heterogeneous cellular networks where the
    base stations in different tiers usually adopt different transmission powers,
    we propose a Transmission Power Normalization Model (TPNM), which is able to
    convert a multi-tier cellular network into a single-tier network, such that all
    base stations have the same normalized transmission power. Then using TPNM, the
    signal and interference received at any point in the complex multi-tier
    environment can be analyzed by considering the same point in the equivalent
    single-tier cellular network model, thus significantly simplifying the
    analysis. On this basis, we propose a new user association scheme in
    heterogeneous cellular networks, where the base station that leads to the
    smallest interference to other co-channel mobile stations is chosen from a set
    of candidate base stations that satisfy the quality-of-service (QoS) constraint
    for an intended mobile station. Numerical results show that the proposed user
    association scheme is able to significantly reduce the down-link interference
    compared with existing schemes while maintaining a reasonably good QoS.

    Complex Network Theoretical Analysis on Information Dissemination over Vehicular Networks

    Jingjing Wang, Chunxiao Jiang, Longxiang Gao, Shui Yu, Zhu Han, Yong Ren
    Subjects: Networking and Internet Architecture (cs.NI); Information Theory (cs.IT)

    How to enhance the communication efficiency and quality on vehicular networks
    is one critical important issue. While with the larger and larger scale of
    vehicular networks in dense cities, the real-world datasets show that the
    vehicular networks essentially belong to the complex network model. Meanwhile,
    the extensive research on complex networks has shown that the complex network
    theory can both provide an accurate network illustration model and further make
    great contributions to the network design, optimization and management. In this
    paper, we start with analyzing characteristics of a taxi GPS dataset and then
    establishing the vehicular-to-infrastructure, vehicle-to-vehicle and the hybrid
    communication model, respectively. Moreover, we propose a clustering algorithm
    for station selection, a traffic allocation optimization model and an
    information source selection model based on the communication performances and
    complex network theory.

    A computational investigation of the relationships between single-neuron and network dynamics in the cerebral cortex

    Stefano Cavallari
    Comments: PhD thesis (integrate-and-fire neurons, recurrent neural networks, current based neurons, conductance based neurons, LFP, EEG, information encoding, GLM, Wiener filtering, PV-pos interneuron) 170 pages, 57 figures, 8 tables
    Subjects: Neurons and Cognition (q-bio.NC); Information Theory (cs.IT); Quantitative Methods (q-bio.QM)

    Functions of brain areas in complex animals are believed to rely on the
    dynamics of networks of neurons rather than on single neurons. On the other
    hand, the network dynamics reflect and arise from the integration and
    coordination of the activity of populations of single neurons. Understanding
    how single-neurons and neural-circuits dynamics complement each other to
    produce brain functions is thus of paramount importance. LFPs and EEGs are good
    indicators of the dynamics of mesoscopic and macroscopic populations of
    neurons, while microscopic-level activities can be documented by measuring the
    membrane potential, the synaptic currents or the spiking activity of individual
    neurons. In this thesis we develop mathematical modelling and mathematical
    analysis tools that can help the interpretation of joint measures of neural
    activity at microscopic and mesoscopic or macroscopic scales. In particular, we
    develop network models of recurrent cortical circuits that can clarify the
    impact of several aspects of single-neuron (i.e., microscopic-level) dynamics
    on the activity of the whole neural population (as measured by LFP). We then
    develop statistical tools to characterize the relationship between the action
    potential firing of single neurons and mass signals. We apply these latter
    analysis techniques to joint recordings of the firing activity of individual
    cell-type identified neurons and mesoscopic (i.e., LFP) and macroscopic (i.e.,
    EEG) signals in the mouse neocortex. We identified several general aspects of
    the relationship between cell-specific neural firing and mass circuit activity,
    providing for example general and robust mathematical rules which infer
    single-neuron firing activity from mass measures such as the LFP and the EEG.

    Unified Quantification of Nonclassicality and Non-Gaussianity: An Entropic Approach

    Soumyakanti Bose
    Comments: 6 pages, 3 figures
    Subjects: Quantum Physics (quant-ph); Information Theory (cs.IT); Optics (physics.optics)

    Nonclassical states of quantized light, except for the Gaussian states, also
    possess non-Gaussian phase-space distributions. Despite several attempts, a
    unified description of nonclassicality (NC) and non-Gaussianity (NG) of quantum
    states of light has not been developed as-of-yet. Here, we propose an
    experimentally verifiable scheme for quantification of NC, in terms of Wehrl
    entropy, that further leads to the simultaneous quantification of NC and NG.
    While requiring no optimization, present work recovers earlier results
    qualitatively as well as explores several new possibilities on the conjugation
    of nonclassical and non-Gaussian character of quantum states. Moreover, current
    formalism, due to its possible extension to the finite-dimensional systems,
    bridges the gap between discrete and continuous variable systems. Our work,
    thus, becomes crucial in describing NC of quantum processes including open
    quantum systems as well as understanding the role of NC and NG as resources in
    several information theoretic tasks processing such as entanglement
    distillation, quantum network, quantum computation etc.

    Finite size analysis of the detectability limit of the stochastic block model

    Jean-Gabriel Young, Patrick Desrosiers, Laurent Hébert-Dufresne, Edward Laurence, Louis J. Dubé
    Comments: Main text: 16 pages, 4 figures. Supplemental Information: 11 pages, 2 figures
    Subjects: Physics and Society (physics.soc-ph); Information Theory (cs.IT)

    It has been shown in recent years that the stochastic block model is
    undetectable in the sparse limit, i.e., that no algorithm can identify a
    partition correlated with the partition used to generate an instance, if the
    instance is sparse and infinitely large. Real networks are however finite
    objects, and one cannot expect all results derived in the infinite limit to
    hold for finite instances. In this contribution, we treat the finite case
    explicitly. We give a necessary condition for finite size detectability in the
    general SBM, using arguments drawn from information theory and statistics. We
    then distinguish the concept of average detectability from the concept of
    instance-by-instance detectability, and give explicit formulas for both
    definitions. Using these formulas, we prove that there exist large equivalence
    classes of parameters, where widely different network ensembles are equally
    detectable with respect to our definitions of detectability. In an extensive
    case study, we investigate the finite size detectability of a simplified
    variant of the SBM, which encompasses a number of important models as special
    cases. These models include the symmetric SBM, the planted coloring model, and
    more exotic SBMs not previously studied. We obtain a number of explicit
    expressions for this variant, and also show that the well-known Kesten-Stigum
    bound does not capture the phenomenon of finite size detectability—even at
    the qualitative level. We conclude with two Appendices, where we study the
    interplay of noise and detectability, and establish a connection between our
    information-theoretic approach and Random Matrix Theory.

    Compressed sensing and optimal denoising of monotone signals

    Eftychios A. Pnevmatikakis
    Comments: To appear in the 42nd IEEE International Conference on Acoustics, Speech and Signal Processing ICASSP2017
    Subjects: Statistics Theory (math.ST); Information Theory (cs.IT)

    We consider the problems of compressed sensing and optimal denoising for
    signals )mathbf{x_0}inmathbb{R}^N( that are monotone, i.e.,
    )mathbf{x_0}(i+1) geq mathbf{x_0}(i)(, and sparsely varying, i.e.,
    )mathbf{x_0}(i+1) > mathbf{x_0}(i)( only for a small number )k( of indices
    )i(. We approach the compressed sensing problem by minimizing the total
    variation norm restricted to the class of monotone signals subject to equality
    constraints obtained from a number of measurements )Amathbf{x_0}(. For random
    Gaussian sensing matrices )Ainmathbb{R}^{m imes N}( we derive a closed form
    expression for the number of measurements )m( required for successful
    reconstruction with high probability. We show that the probability undergoes a
    phase transition as )m( varies, and depends not only on the number of change
    points, but also on their location. For denoising we regularize with the same
    norm and derive a formula for the optimal regularizer weight that depends only
    mildly on )mathbf{x_0}(. We obtain our results using the statistical dimension
    tool.




沪ICP备19023445号-2号
友情链接