Physical Sciences and Mathematics Commons^™

Computational Communication Intelligence: Exploring Linguistic Manifestation And Social Dynamics In Online Communication, Xiaoxi Xu

Doctoral Dissertations

We now live in an age of online communication. As social media becomes an integral part of our life, online communication becomes an essential life skill. In this dissertation, we aim to understand how people effectively communicate online. We research components of success in online communication and present scientific methods to study the skill of effective communication. This research advances the state of art in machine learning and communication studies. For communication studies, we pioneer the study of a communication phenomenon we call Communication Intelligence in online interactions. We create a theory about communication intelligence that measures participants’ ten high-order …

Data Analysis And Study Design In The Presence Of Error-Prone Diagnostic Tests, Xiangdong Gu

Doctoral Dissertations

Interval censored time to event outcomes arise when a silent event of interest is known to have occurred within a specific time period, determined by the times of the last negative and first positive diagnostic tests. The four chapters comprising this thesis are tied together by a common theme in that the outcome of interest is an interval censored time to event random variable. In Chapter 1, we describe a stratified Weibull model appropriate for interval cen- sored outcomes and implement a new R package straweib. We compare the proposed approach with the log-linear form of the Weibull regression model …

Scaling Mcmc Inference And Belief Propagation To Large, Dense Graphical Models, Sameer Singh

Doctoral Dissertations

With the physical constraints of semiconductor-based electronics becoming increasingly limiting in the past decade, single-core CPUs have given way to multi-core and distributed computing platforms. At the same time, access to large data collections is progressively becoming commonplace due to the lowering cost of storage and bandwidth. Traditional machine learning paradigms that have been designed to operate sequentially on single processor architectures seem destined to become obsolete in this world of multi-core, multi-node systems and massive data sets. Inference for graphical models is one such example for which most existing algorithms are sequential in nature and are difficult to scale …

Evaluating Predictors Of An Individual’S Dietary Intake Latent Value Under Different Mixed Models, Shuli Yu

Doctoral Dissertations

The accurate estimation of an individual’s usual dietary intake is important since the estimates are essential to uncover the diet-disease relationships. This study explores a more accurate method to estimate an individual’s latent value of usual dietary intake when it is repeatedly measured using a 24-hour dietary recall (24HR) and seven day dietary recall (7DDR), accounting for random measurement error and bias. The performance of the (empirical) predictor of subject’s latent value obtained under the finite population mixed model (FPMM) framework is compared with those obtained under the usual mixed model and the measurement error model through a simulation study. …

Incorporating Boltzmann Machine Priors For Semantic Labeling In Images And Videos, Andrew Kae

Doctoral Dissertations

Semantic labeling is the task of assigning category labels to regions in an image. For example, a scene may consist of regions corresponding to categories such as sky, water, and ground, or parts of a face such as eyes, nose, and mouth. Semantic labeling is an important mid-level vision task for grouping and organizing image regions into coherent parts. Labeling these regions allows us to better understand the scene itself as well as properties of the objects in the scene, such as their parts, location, and interaction within the scene. Typical approaches for this task include the conditional random field …

Predicting High-Stakes Tests Of Math Achievement Using A Group-Administered Rti Instrument: Validating Skills Measured By The Monitoring Instructional Responsiveness: Math, Jeremy Thomas Coles

Doctoral Dissertations

Three universal screeners and nine progress monitoring probes from the Monitoring Instructional Responsiveness: Math (MIR:M), a silent, group-administered math assessment designed for implementation with an RTI Model, were administered to 223 fifth-grade students. The growth parameters of the overall MIR:M composite and two global composites (math calculation and math reasoning) identified significant variation in student growth, within significant linear and quadratic trajectories. However, there were significant differences in the nature of the growth trajectories that have applied educational implications. In addition, growth parameters across the three composites provided significant predictive potential when using the Tennessee Comprehensive Assessment Program (TCAP) Achievement …

Impacts Of Climate Change On The Evolution Of The Electrical Grid, Melissa Ree Allen

Doctoral Dissertations

Maintaining interdependent infrastructures exposed to a changing climate requires understanding 1) the local impact on power assets; 2) how the infrastructure will evolve as the demand for infrastructure changes location and volume and; 3) what vulnerabilities are introduced by these changing infrastructure topologies. This dissertation attempts to develop a methodology that will a) downscale the climate direct effect on the infrastructure; b) allow population to redistribute in response to increasing extreme events that will increase under climate impacts; and c) project new distributions of electricity demand in the mid-21^st century.

The research was structured in three parts. The first …

Common Method Variance: An Experimental Manipulation, Alison Wall

Doctoral Dissertations

Although common method variance has been a subject of research concern for over fifty years, its influence on study results is still not well understood. Common method variance concerns are frequently cited as an issue in the publication of self-report data; yet, there is no consensus as to when, or if, common method variance creates bias. This dissertation examines common method variance by approaching it from an experimental standpoint. If groups of respondents can be influenced to vary their answers to survey items based upon the presence or absence of procedural remedies, a better understanding of common method variance can …

Improvements On Segment Based Contours Method For Dna Microarray Image Segmentation, Yang Li

Doctoral Dissertations

DNA microarray is an efficient biotechnology tool for scientists to measure the expression levels of large numbers of genes, simultaneously. To obtain the gene expression, microarray image analysis needs to be conducted. Microarray image segmentation is a fundamental step in the microarray analysis process. Segmentation gives the intensities of each probe spot in the array image, and those intensities are used to calculate the gene expression in subsequent analysis procedures. Therefore, more accurate and efficient microarray image segmentation methods are being pursued all the time.

In this dissertation, we are making efforts to obtain more accurate image segmentation results. We …

Associations Of Total Activity Counts And Physical Activity Intensity Levels With The Metabolic Syndrome: A Structural Equation Modeling Approach, Dana Lizbeth Wolff

Doctoral Dissertations

To clarify the protective benefits of physical activity (PA), epidemiologists and public health researchers continue to seek improved methods of assessing PA. In particular, accelerometers have gained acceptance with researchers as they provide reliable estimates of PA and can record both the amount and intensity of ambulatory movement. However, there is concern that accelerometer data reduction techniques may not provide quantitatively accurate measurements of time spent in various PA intensity categories. One way to circumvent these inaccuracies is to use the accelerometer-derived total activity counts (TAC), which is a more direct expression of what the monitor records.

In order to …

Asymptotic Behavior Of A Class Of Spdes, Parisa Fatheddin

Doctoral Dissertations

We establish the large and moderate deviation principles for a class of stochastic partial differential equations with a non-Lipschitz continuous coefficient. As an application we derive these principles for an important population model, Fleming-Viot Process. In addition, we establish the moderate deviation principle for another classical population model, super-Brownian motion.

Performance Modeling And Optimization Techniques For Heterogeneous Computing, Supada Laosooksathit

Doctoral Dissertations

Since Graphics Processing Units (CPUs) have increasingly gained popularity amoung non-graphic and computational applications, known as General-Purpose computation on GPU (GPGPU), CPUs have been deployed in many clusters, including the world's fastest supercomputer. However, to make the most efficiency from a GPU system, one should consider both performance and reliability of the system.

This dissertation makes four major contributions. First, the two-level checkpoint/restart protocol that aims to reduce the checkpoint and recovery costs with a latency hiding strategy in a system between a CPU (Central Processing Unit) and a GPU is proposed. The experimental results and analysis reveals some benefits, …

Vulnerability Analysis Of Cyber-Behavioral Biometric Authentication, Abdul Serwadda

Doctoral Dissertations

Research on cyber-behavioral biometric authentication has traditionally assumed naïve (or zero-effort) impostors who make no attempt to generate sophisticated forgeries of biometric samples. Given the plethora of adversarial technologies on the Internet, it is questionable as to whether the zero-effort threat model provides a realistic estimate of how these authentication systems would perform in the wake of adversity. To better evaluate the efficiency of these authentication systems, there is need for research on algorithmic attacks which simulate the state-of-the-art threats.

To tackle this problem, we took the case of keystroke and touch-based authentication and developed a new family of algorithmic …