Title: Statistical Disclosure Attacks on Anonymous Communication Systems
Abstract: In this talk, I will introduce the problem of how to provide anonymity, i.e., hide who communicates with whom, in a communications system. We will focus on mixes, the main building blocks of high-latency anonymous communication channels, explaining how they work in order to hide the correspondences between the senders and receivers of messages, and we will overview the basic attacks against these devices. Finally, we will see how to optimize the design of the system in order to achieve better privacy guarantees using signal processing tools.
Biography: Simon Oya received the Telecommunication Engineer degree from the University of Vigo, Spain, in 2012, and the master’s degree in Signal Processing in Communications, in 2014. He is pursuing the Ph.D. degree in Telecommunication Engineering in the same university, and he is currently at Rutgers University as a visitor student under the supervision of Prof. Anand Sarwate. His research interest is the study of privacy-preserving technologies from a signal processing point of view.
Title: Free-Form Gesture Authentication in the Wild
Abstract: Free-form gesture passwords have been introduced as an alternative mobile authentication method. Text passwords are not very suitable for mobile interaction, and methods such as PINs and grid patterns sacrifice security over usability. However, little is known about how free-form gestures perform in the wild. We present the first field study (N=91) of mobile authentication using free-form gestures, with text passwords as a baseline. Our study leveraged Experience Sampling Methodology to increase ecological validity while maintaining control of the experiment. We found that, with gesture passwords, participants generated new passwords and authenticated faster with comparable memorability while being more willing to retry. Our analysis of the gesture password dataset indicated biases in user-chosen distribution tending towards common shapes. Our findings provide useful insights towards understanding mobile device authentication and gesture-based authentication. This paper will be formally published in CHI’16, the ACM annual premier conference in HCI, held in San Jose, CA May 2016.
Biography: Yulong Yang is a fifth-year Ph.D. student in the Department of Electrical and Computer Engineering. His research advisor is Professor Janne Lindqvist. His interests are in usable security, mobile applications and human computer interaction. He has worked on publications that accepted to conferences including Mobisys, Ubicomp, LASER, etc. He received B.S. degree in Electrical Engineering from University of Electronic and Scientific Technology of China in 2011.
Title: Of Two Minds, Multiple Addresses, and One Ledger: Characterizing Opinions, Knowledge, and Perceptions of Bitcoin Across Users and Non-Users
Abstract: Digital currencies represent a new method for exchange – a payment method with no physical form, made real by the Internet. This new type of currency was created to ease online transactions and to provide greater convenience in making payments. However, a critical component of a monetary system is the people who use it. Acknowledging this, we present results of our interview study (N=20) with two groups of participants (users and non-users) about how they perceive the most popular digital currency, Bitcoin. Our results reveal: non-users mistakenly believe they are incapable of using Bitcoin, users are not well-versed in how the protocol functions, they have misconceptions about the privacy of transactions, and that Bitcoin satisfies properties of ideal payment systems as defined by our participants. Our results illustrate Bitcoin’s tradeoffs, its uses, and barriers to entry. The work has been featured in over 100 popular media and web sites so far, including Yahoo! Finance News, Morningstar, IBS Intelligence, NSF’s website, Benzinga, and Neowin. The paper will be formally published in CHI’16, the ACM annual premier conference in HCI, held in San Jose, CA May 2016.
Biography: Xianyi Gao is currently a PhD student in the Department of Electrical and Computer Engineering (ECE) at Rutgers University. He is a recipient of the NSF GRFP award. He received the Best Paper Nominee Award from UbiComp’14. He is currently working under supervision of Dr. Janne Lindqvist on areas related to Human-Computer Interaction. His research interests include security and privacy for cryptocurrencies, automobiles, and other mobile systems. He joined the ECE PhD program in Fall 2013. Before that, he obtained his undergraduate degree in electrical engineering from Rutgers University as well.
Abstract: We examine lossless data compression from an average delay perspective. An encoder receives input symbols one per unit time from an i.i.d. source and submits binary codewords to a FIFO buffer that transmits bits at a fixed rate to a receiver/decoder. Each input symbol at the encoder is viewed as a status update by the source and the system performance is characterized by the status update age, defined as the number of time units (symbols) the decoder output lags behind the encoder input. For lossless block coding schemes, an upper bound on the average status age is derived from error exponent with delay in streaming block coding, but the error exponent does not lead to an accurate description of the status age for small delay and small blocklength. An age optimal block coding scheme is proposed based on an approximation of the average age by converting the streaming source coding system into a D/G/1 queue. We compare this scheme to the error exponent optimal coding scheme which uses the method of types and show that maximizing the error exponent is not equivalent to minimizing the average status age. We also apply the status age analysis to streaming arithmetic coding and show that arithmetic coding provides lower average age at low channel rate since the redundancy converges to zero for long source sequences.
Biography: Jing Zhong is a Ph.D. student in the Department of Electrical and Computer Engineering. He is currently working under the supervision of Prof. Roy Yates on status updating systems and its application to data compression. His research interests include information theory, communication networks and statistical signal processing. He received B.Eng. degree in Electrical Engineering from Hong Kong University of Science and Technology (HKUST) in 2013.
Title: Nonparametric Graphical Model: Foundation and Trends
Abstract: We consider the problem of learning the structure of a non-Gaussian graphical model. We introduce two strategies for constructing tractable nonparametric graphical model families. One approach is through semiparametric extension of the Gaussian or exponential family graphical models that allows arbitrary graphs. Another approach is to restrict the family of allowed graphs to be acyclic, enabling the use of fully nonparametric density estimation in high dimensions. These two approaches can both be viewed as adding structural regularization to the a general pairwise nonparametric Markov random field and reflect an interesting tradeoff of model flexibility with structural complexity. In terms of graph estimation, these methods achieve the optimal parametric rates of convergence. In terms of computation, these methods are as scalable as the best implemented parametric methods. Such a “free lunch phenomenon” make them extremely attractive for large-scale applications. We will also introduce several new research directions along this line of work, including latent-variable extension, model-based nonconvex optimization, graph uncertainty assessment, and nonparametric graph property testing.
Biography: Han Liu is an Assistant Professor in the Department of Operations Research and Financial Engineering at Princeton University, where he leads the Statistical Machine Learning (SMiLe) Laboratory. In 2011, he received the joint PhD in Statistics and Machine Learning from the Machine Learning Department at the Carnegie Mellon University. His theoretical research focuses on nonparametric graphical models, nonconvex statistical optimization, and post-regularization inference. His applied research focuses on brain science, genomics, and computational finance. Han Liu is the recipient of several research awards including the Tweedie New Researcher Award (from IMS), the Noether Young Scholar Award (from ASA), the NSF CAREER Award (from DMS), the Howard B Wentz Award (from Princeton SEAS), and has received several best paper awards in the International Conference on Machine Learning and International Conference on Artificial Intelligence and Statistics.
Title: Radar for Human Activity Classification in Assisted Living Applications
Abstract: The elderly population aged 65+ years is growing and their ratio to the population aged 15-64 is expected to reach 40% by 2030. This implies that those of working age, and, subsequently, the overall economy, will face a greater burden in supporting the aging population. In addition, the demand and trend is upward for continued independent living. As such, there is a growing interest in assisted living technologies that enable self-dependent living within homes and residences for the elderly. Remote monitoring capabilities, such as detection of falls and small changes in motor functional abilities of the elderly, will address the challenges associated with self-dependent living. This talk focuses on the radar technology and discusses the time-frequency based nonstationary signal processing techniques used to provide the local signal behavior over frequency and to detail the changes in the Doppler and micro-Doppler radar signatures over time. Features that capture the intrinsic differences in the time-frequency signatures of different gross motor activities of the elderly are identified and their performance in human activity classification is demonstrated using real data measurements. Offerings of the range information, in addition to Doppler, for classifying different motion articulations with enhanced reliability are also highlighted.
Biography: Dr. Fauzia Ahmad received her Ph.D. in Electrical Engineering from the University of Pennsylvania in 1997. Currently, she is a Research Professor and the Director of the Radar Imaging Laboratory at the Center for Advanced Communications, College of Engineering, Villanova University. She is a Senior Member of the IEEE and SPIE. She is an Associate Editor of the IEEE Transactions on Signal Processing and IEEE Geoscience and Remote Sensing Letters, and is a member of the editorial board of the IET Radar Sonar & Navigation and SPIE/IS&T Journal of Electronic Imaging. Dr. Ahmad is a member of the Radar Systems Panel of the IEEE Aerospace and Electronic Systems Society, Sensor Array and Multichannel Technical Committee of the IEEE Signal Processing Society, and the Electrical Cluster of the Franklin Institute Committee on Science and the Arts. She has also been the Chair of the SPIE Compressive Sensing Conference Series since 2012. Dr. Ahmad has 190+ journal and conference publications and seven book chapters in the areas of radar imaging, radar signal processing, array signal processing, waveform design, compressive sensing, detection and localization, direction finding, and ultrasound imaging. She has been a PI/Co-PI on various projects in the aforementioned areas with the total research awards exceeding $6.5M.
Title: From Seismology to Compressed Sensing and Back, a Brief History of Optimization-Based Signal Processing
Abstract: In this talk we provide an overview of the history of l1-norm minimization applied to underdetermined inverse problems. In the 70s and 80s geophysicists proposed using l1-norm minimization for deconvolution from bandpass data in reflection seismography. In the 2000s, inspired by this approach and by magnetic resonance imaging, a method to provably recover sparse signals from random projections, known as compressed sensing, was developed. Theoretical insights used to analyze compressed sensing have recently been adapted to understand the potential and limitations of l1-norm minimization for deterministic problems. These include super-resolution from low-pass data and the deconvolution problem that originally motivated the geophysicists.
Biography: Carlos Fernandez-Granda received engineering degrees from Universidad Politecnica de Madrid and Ecole des Mines in Paris and an M. Sc. from Ecole Normale Superieure de Cachan. For his master’s thesis he investigated image reconstruction in parallel magnetic resonance imaging at Philips Research. He then obtained a Ph. D. from Stanford University, where he studied the problem of super-resolving signals from blurred data using methods based on convex optimization. Before joining Courant, he spent a year at Google, where he worked on techniques to process neural data. His research focuses on developing and analyzing optimization-based methods to tackle inverse problems that arise in applications such as neuroscience, computer vision and medical imaging.
Title: Bilinear Regression with Matrix Covariates and Applications in Neuroimaging Data Analysis
Abstract: Traditional functional linear regression usually takes a one dimensional functional predictor as input and estimates the continuous coefficient function. Modern applications often generate two dimensional covariates, which when observed at grid points are matrices. To avoid inefficiency of the classical method involving estimation of a two dimensional coefficient function, we propose a bilinear regression model and obtain estimates via a smoothness regularization method. The proposed estimator exhibits minimax optimal property for prediction under the framework of Reproducing Kernel Hilbert Space. The merits of the method are further demonstrated by numerical experiments and an application on real imaging data.
Biography: Dan Yang has a PhD in Statistics from University of Pennsylvania, and an undergraduate degree in Statistics from Peking University. She is currently an assistant professor in the Department of Statistics and Biostatistics at Rutgers University. Her research interests include high dimensional data analysis, functional data analysis, multivariate data analysis, nonparametric statistics, analysis of observational studies, and imaging data.
Title: Quickest Detection and Isolation of Line Outages in Power Systems
Abstract: The problem of detecting abrupt changes in stochastic systems and time series, often referred to as the quickest change detection (QCD) problem, arises in various branches of science and engineering. It is assumed that the observations undergo achange in distribution in response to a change or disruption in the environment or, more generally, to changes in certain patterns. The observations are obtained sequentially and, as long as their behavior is consistent with the normal state, the process is allowed to continue. If the state changes, then it is of interest to detect the change as soon as possible, subject to false alarm constraints, and take any necessary action in response to the change. In the first part of this talk, an up-to-date overview of the results on the QCD problem will be provided, including some recent results on data-efficient QCD. A number of applications of QCD will be discussed. In the second part of the talk, the focus will be on the problem of detecting and isolating line outages in power systems using phasor measurement unit (PMU) measurements taken at the buses. It is shown that QCD based algorithms are tailor-made for this problem and significantly outperform existing methods.
Biography: Dr. Veeravalli, received the Ph.D. degree in Electrical Engineering from the University of Illinois at Urbana-Champaign in 1992, the M.S. degree from Carnegie-Mellon University in 1987, and the B.Techdegree from Indian Institute of Technology, Bombay (Silver Medal Honors) in 1985. He is currently a Professor in the department of Electrical and Computer Engineering (ECE), the Coordinated Science Laboratory (CSL) and the Information Trust Institute (ITI) at the University of Illinois at Urbana-Champaign. He was on the faculty of the School of ECE at Cornell University before he joined Illinois in 2000. He served as a program director for communications research at the U.S. National Science Foundation in Arlington, VA during 2003-2005. His research interests span the theoretical areas of detection and estimation, information theory, statistical learning, and stochastic control, with applications to wireless communication systems and networks, sensor networks, cyberphysicalsystems, big data, and genomics. He is a Fellow of the IEEE, and a recipient of the 1996 IEEE Browder J. Thompson Best Paper Award and the U.S. Presidential Early Career Award for Scientists and Engineers (PECASE). He served as a distinguished lecturer for the IEEE Signal Processing Society during 2010-2011.
Title: Distributed Proportional Stochastic Coordinate Descent With Social Sampling
Abstract: We consider stochastic message passing algorithms that limit the communication required for decentralized and distributed convex optimization and provide convergence guarantees on the objective value. We first propose a centralized method that modifies the coordinate-sampling distribution for stochastic coordinate descent, which we call proportional stochastic coordinate descent. This method treats the gradient of the function as a probability distribution to sample the coordinates, and may be useful in so-called lock-free decentralized optimization schemes. For general distributed optimization in which agents jointly minimize the sum of local objectives we propose treating the iterates as gradients and propose a stochastic coordinate-wise primal averaging algorithm for optimization.
Biography: Mohsen Ghassemi has an undergraduate degree in electrical engineering from University of Tehran, Iran. He joined Rutgers in fall 2013 as a PhD student. He is currently working under supervision of Prof. Anand Sarwate on stochastic and distributed optimization algorithms for large scale machine learning problems.
Title: Symmetric Matrix Perturbation For Differentially-Private Principal Component Analysis
Abstract: Differential privacy is a strong, cryptographically-motivated definition of privacy that has recently received a significant amount of research attention for its robustness to known attacks. The principal component analysis (PCA) algorithm is frequently used in signal processing, machine learning, statistics pipelines. In this paper, we propose a new algorithm for differentially-private computation of PCA and compare the performance empirically with some recent state-of-the-art algorithms on different data sets. We intend to investigate the performance of these algorithms with varying privacy parameters and database parameters. We show that our proposed algorithm, despite guaranteeing stricter privacy, provides very good utility for different data sets.
Biography: Hafiz Imtiaz earned his B.Sc. and M.Sc. degrees in Electrical and Electronic Engineering from Bangladesh University of Engineering and Technology, Dhaka, Bangladesh. He joined Rutgers in Fall 2014 as a Ph.D. student under the supervision of Prof. Anand Sarwate. He is currently working on differential privacy and related applications.