Numerical Analysis and Scientific Computing Commons^™

Generative Modeling Of Entity Comparisons In Text, Maksim Tkachenko, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Users frequently rely on online reviews for decision making. In addition to allowing users to evaluate the quality of individual products, reviews also support comparison shopping. One key user activity is to compare two (or more) products based on a specific aspect. However, making a comparison across two different reviews, written by different authors, is not always equitable due to the different standards and preferences of individual authors. Therefore, we focus instead on comparative sentences, whereby two products are compared directly by a review author within a single sentence. We study the problem of comparative relation mining. Given a set …

Online Passive Aggressive Active Learning And Its Applications, Jing Lu, Peilin Zhao, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

We investigate online active learning techniques for classification tasks in data stream mining applications. Unlike traditional learning approaches (either batch or online learning) that often require to request the class label of each incoming instance, online active learning queries only a subset of informative incoming instances to update the classification model, which aims to maximize classification performance using minimal human labeling effort during the entire online stream data mining task. In this paper, we present a new family of algorithms for online active learning called Passive-Aggressive Active (PAA) learning algorithms by adapting the popular Passive-Aggressive algorithms in an online active …

Networked Employment Discrimination, Tamara Kneese

Media Studies

Employers often struggle to assess qualified applicants, particularly in contexts where they receive hundreds of applications for job openings. In an effort to increase efficiency and improve the process, many have begun employing new tools to sift through these applications, looking for signals that a candidate is “the best fit.” Some companies use tools that offer algorithmic assessments of workforce data to identify the variables that lead to stronger employee performance, or to high employee attrition rates, while others turn to third party ranking services to identify the top applicants in a labor pool. Still others eschew automated systems, but …

Partisan Sharing: Facebook Evidence And Societal Consequences, Jisun An, Daniele Quercia, Jon Crowcroft

Research Collection School Of Computing and Information Systems

The hypothesis of selective exposure assumes that people seek out information that supports their views and eschew information that conflicts with their beliefs, and that has negative consequences on our society. Few researchers have recently found counter evidence of selective exposure in social media: users are exposed to politically diverse articles. No work has looked at what happens after exposure, particularly how individuals react to such exposure, though. Users might well be exposed to diverse articles but share only the partisan ones. To test this, we study partisan sharing on Facebook: the tendency for users to predominantly share like-minded news …

Cost-Sensitive Online Classification, Jialei Wang, Peilin Zhao, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Both cost-sensitive classification and online learning have been extensively studied in data mining and machine learning communities, respectively. However, very limited study addresses an important intersecting problem, that is, “Cost-Sensitive Online Classification". In this paper, we formally study this problem, and propose a new framework for Cost-Sensitive Online Classification by directly optimizing cost-sensitive measures using online gradient descent techniques. Specifically, we propose two novel cost-sensitive online classification algorithms, which are designed to directly optimize two well-known cost-sensitive measures: (i) maximization of weighted sum of sensitivity and specificity, and (ii) minimization of weighted misclassification cost. We analyze the theoretical bounds of …

Numerical Studies Of The A6 Relativistic Magnetron Using Particle-In-Cell Simulations, Vineetha Jujjavarapu

Electrical & Computer Engineering Theses & Dissertations

The magnetron became a major component in radar systems around the time of the World War II. More recent radar techniques require short microwave pulses (with nanosecond pulse-widths for ultra wideband operations) and high peak power in the Giga-Watt range. In order to achieve these specifications for the military, the magnetron needs to have both a fast start and the rapid growth rate of oscillations in order to serve as a microwave source for the short-pulse applications. These two factors depend on the geometry of the cathode. Other requirements include high conversion efficiency and low leakage current (in order to …

Enhanced Rare-Region Effects In The Contact Process With Long-Range Correlated Disorder, Ahmed K. Ibrahim, Hatem Barghathi, Thomas Vojta

Physics Faculty Research & Creative Works

We investigate the nonequilibrium phase transition in the disordered contact process in the presence of long-range spatial disorder correlations. These correlations greatly increase the probability for finding rare regions that are locally in the active phase while the bulk system is still in the inactive phase. Specifically, if the correlations decay as a power of the distance, the rare-region probability is a stretched exponential of the rare-region size rather than a simple exponential as is the case for uncorrelated disorder. As a result, the Griffiths singularities are enhanced and take a non-power-law form. The critical point itself is of infinite-randomness …

High Performance Computing Markov Models Using Hadoop Mapreduce, Matthew Shaffer

e-Research: A Journal of Undergraduate Work

In this paper, I will explain how I used the probability modeling tool, Markov Models, in combination with Hadoop MapReduce parallel programming platform in order to quickly and efficiently analyses documents and create a probability model of them. I will explain what Markov Models are, give a brief overview of what MapReduce is, explain why Markov models can be used for document analysis, explain my code of the modeling program, and examine the performance of various MapReduce platforms and techniques in analyzing documents.

Sharing Political News: The Balancing Act Of Intimacy And Socialization In Selective Exposure, Jisun An, Daniele Quercia, Meeyoung Cha, Krishna Gummadi, Jon Crowcroft

Research Collection School Of Computing and Information Systems

One might think that, compared to traditional media, social media sites allow people to choose more freely what to read and what to share, especially for politically oriented news. However, reading and sharing habits originate from deeply ingrained behaviors that might be hard to change. To test the extent to which this is true, we propose a Political News Sharing (PoNS) model that holistically captures four key aspects of social psychology: gratification, selective exposure, socialization, and trust & intimacy. Using real instances of political news sharing in Twitter, we study the predictive power of these features. As one might expect, …

Seeing Human Weight From A Single Rgb-D Image, Tam Nguyen, Jiashi Feng, Shuicheng Yan

Computer Science Faculty Publications

Human weight estimation is useful in a variety of potential applications, e.g., targeted advertisement, entertainment scenarios and forensic science. However, estimating weight only from color cues is particularly challenging since these cues are quite sensitive to lighting and imaging conditions. In this article, we propose a novel weight estimator based on a single RGB-D image, which utilizes the visual color cues and depth information. Our main contributions are three-fold.

First, we construct the W8-RGBD dataset including RGB-D images of different people with ground truth weight.

Second, the novel sideview shape feature and the feature fusion model are proposed to facilitate …

Online Probabilistic Learning For Fuzzy Inference System, Richard Jayadi Oentaryo, Meng Joo Er, San Linn, Xiang Li

Research Collection School Of Computing and Information Systems

Online learning is a key methodology for expert systems to gracefully cope with dynamic environments. In the context of neuro-fuzzy systems, research efforts have been directed toward developing online learning methods that can update both system structure and parameters on the fly. However, the current online learning approaches often rely on heuristic methods that lack a formal statistical basis and exhibit limited scalability in the face of large data stream. In light of these issues, we develop a new Sequential Probabilistic Learning for Adaptive Fuzzy Inference System (SPLAFIS) that synergizes the Bayesian Adaptive Resonance Theory (BART) and Rule-Wise Decoupled Extended …

Interestingness-Driven Diffussion Process Summarization In Dynamic Networks, Qiang Qu, Siyuan Liu, Christian Jensen, Feida Zhu, Christos Faloutsos

Research Collection School Of Computing and Information Systems

The widespread use of social networks enables the rapid diffusion of information, e.g., news, among users in very large communities. It is a substantial challenge to be able to observe and understand such diffusion processes, which may be modeled as networks that are both large and dynamic. A key tool in this regard is data summarization. However, few existing studies aim to summarize graphs/networks for dynamics. Dynamic networks raise new challenges not found in static settings, including time sensitivity and the needs for online interestingness evaluation and summary traceability, which render existing techniques inapplicable. We study the topic of dynamic …

Phase Transitions On Random Lattices: How Random Is Topological Disorder?, Hatem Barghathi, Thomas Vojta

Physics Faculty Research & Creative Works

We study the effects of topological (connectivity) disorder on phase transitions. We identify a broad class of random lattices whose disorder fluctuations decay much faster with increasing length scale than those of generic random systems, yielding a wandering exponent of ω = (d−1)/(2d) in d dimensions. The stability of clean critical points is thus governed by the criterion (d+1)ν > 2 rather than the usual Harris criterion dν > 2, making topological disorder less relevant than generic randomness. The Imry-Ma criterion is also modified, allowing first-order transitions to survive in all dimensions d > 1. These results explain a host of puzzling violations …

Strong-Disorder Magnetic Quantum Phase Transitions: Status And New Developments, Thomas Vojta

Physics Faculty Research & Creative Works

This article reviews the unconventional effects of random disorder on magnetic quantum phase transitions, focusing on a number of new experimental and theoretical developments during the last three years. On the theory side, we address smeared quantum phase transitions tuned by changing the chemical composition, for example in alloys of the type A_1-xB_x. We also discuss how the interplay of order parameter conservation and overdamped dynamics leads to enhanced quantum Griffiths singularities in disordered metallic ferromagnets. Finally, we discuss a semiclassical theory of transport properties in quantum Griffiths phases. Experimental examples include the ruthenates Sr_{1-x …}

Granular Matter: Microstructural Evolution And Mechanical Response, Aashish Ghimire, Ishan Srivastava, Timothy S. Fisher

The Summer Undergraduate Research Fellowship (SURF) Symposium

Heterogeneous (nano) composites, manufactured by the densification of variously sized grains, represent an important and ubiquitous class of technologically relevant materials. Typical grain sizes in such materials range from macroscopic to a few nanometers. The morphology exhibited by such disordered materials is complex and intricately connected with its thermal and electrical transport properties. It is important to quantify the geometric features of these materials and simulate the fabrication process. Additionally, granular materials exhibit complex structural and mechanical properties that crucially govern their reliability during industrial use. In this work, we simulate the densification of soft deformable grains from a low-density …

Bayesian Calibration Tool, Sveinn Palsson, Martin Hunt, Alejandro Strachan

The Summer Undergraduate Research Fellowship (SURF) Symposium

Fitting a model to data is common practice in many fields of science. The models may contain unknown parameters and often, the goal is to obtain good estimates of them. A variety of methods have been developed for this purpose. They often differ in complexity, efficiency and accuracy and some may have very limited applications. Bayesian inference methods have recently become popular for the purpose of calibrating model's parameters. The way they treat unknown quantities is completely different from any classical methods. Even though the unknown quantity is a constant, it is treated as a random variable and the desired …

Opinion Mining Of Sociopolitical Comments From Social Media, Swapna Gottipati

Dissertations and Theses Collection (Open Access)

Opinions are central to almost all human activities by influencing greatly the decision making process. In this thesis, we present the problems of mining issues, extracting entities and suggestive opinions towards the entities, detecting thoughtful comments, and extracting stances and ideological expressions from online comments in the sociopolitical domain. This study is essential for opinion mining applications that are beneficial for policy makers, government sectors and social organizations. Much work has been done to try to uncover consumer sentiments from online comments to help businesses improve their products and services. However, sociopolitical opinion mining poses new challenges due to complex …

A Mathematical Model And Metaheuristics For Time Dependent Orienteering Problem, Aldy Gunawan, Zhi Yuan, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

This paper presents a generalization of the Orienteering Problem, the Time-Dependent Orienteering Problem (TDOP) which is based on the real-life application of providing automatic tour guidance to a large leisure facility such as a theme park. In this problem, the travel time between two nodes depends on the time when the trip starts. We formulate the problem as an integer linear programming (ILP) model. We then develop various heuristics in a step by step fashion: greedy construction, local search and variable neighborhood descent, and two versions of iterated local search. The proposed metaheuristics were tested on modified benchmark instances, randomly …

Semantic Visualization For Spherical Representation, Tuan M. V. Le, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Visualization of high-dimensional data such as text documents is widely applicable. The traditional means is to find an appropriate embedding of the high-dimensional representation in a low-dimensional visualizable space. As topic modeling is a useful form of dimensionality reduction that preserves the semantics in documents, recent approaches aim for a visualization that is consistent with both the original word space, as well as the semantic topic space. In this paper, we address the semantic visualization problem. Given a corpus of documents, the objective is to simultaneously learn the topic distributions as well as the visualization coordinates of documents. We propose …

Interpretable Machine Learning And Sparse Coding For Computer Vision, Will Landecker

Dissertations and Theses

Machine learning offers many powerful tools for prediction. One of these tools, the binary classifier, is often considered a black box. Although its predictions may be accurate, we might never know why the classifier made a particular prediction. In the first half of this dissertation, I review the state of the art of interpretable methods (methods for explaining why); after noting where the existing methods fall short, I propose a new method for a particular type of black box called additive networks. I offer a proof of trustworthiness for this new method (meaning a proof that my method does not …