Physical Sciences and Mathematics Commons^™

Feature-Based Transfer Learning In Natural Language Processing, Jianfei Yu

Dissertations and Theses Collection (Open Access)

In the past few decades, supervised machine learning approach is one of the most important methodologies in the Natural Language Processing (NLP) community. Although various kinds of supervised learning methods have been proposed to obtain the state-of-the-art performance across most NLP tasks, the bottleneck of them lies in the heavy reliance on the large amount of manually annotated data, which is not always available in our desired target domain/task. To alleviate the data sparsity issue in the target domain/task, an attractive solution is to find sufficient labeled data from a related source domain/task. However, for most NLP applications, due to …

Reinforcement Learning For Collective Multi-Agent Decision Making, Duc Thien Nguyen

Dissertations and Theses Collection (Open Access)

In this thesis, we study reinforcement learning algorithms to collectively optimize decentralized policy in a large population of autonomous agents. We notice one of the main bottlenecks in large multi-agent system is the size of the joint trajectory of agents which quickly increases with the number of participating agents. Furthermore, the noiseof actions concurrently executed by different agents in a large system makes it difficult for each agent to estimate the value of its own actions, which is well-known as the multi-agent credit assignment problem. We propose a compact representation for multi-agent systems using the aggregate counts to address …

Modeling Movement Decisions In Networks: A Discrete Choice Model Approach, Larry Lin Junjie

Dissertations and Theses Collection (Open Access)

In this dissertation, we address the subject of modeling and simulation of agents and their movement decision in a network environment. We emphasize the development of high quality agent-based simulation models as a prerequisite before utilization of the model as an evaluation tool for various recommender systems and policies. To achieve this, we propose a methodological framework for development of agent-based models, combining approaches such as discrete choice models and data-driven modeling.

The discrete choice model is widely used in the field of transportation, with a distinct utility function (e.g., demand or revenue-driven). Through discrete choice models, the movement decision …

Empathetic Computing For Inclusive Application Design, Kenny Choo Tsu Wei

Dissertations and Theses Collection (Open Access)

The explosive growth of the ecosystem of personal and ambient computing de- vices coupled with the proliferation of high-speed connectivity has enabled ex- tremely powerful and varied mobile computing applications that are used every- where. While such applications have tremendous potential to improve the lives of impaired users, most mobile applications have impoverished designs to be inclusive– lacking support for users with specific disabilities. Mobile app designers today haveinadequate support to design existing classes of apps to support users with specific disabilities, and more so, lack the support to design apps that specifically target these users. One way to resolve …

Sensor-Based Human Activity Recognition Using Bidirectional Lstm For Closely Related Activities, Arumugam Thendramil Pavai

Electronic Theses, Projects, and Dissertations

Recognizing human activities using deep learning methods has significance in many fields such as sports, motion tracking, surveillance, healthcare and robotics. Inertial sensors comprising of accelerometers and gyroscopes are commonly used for sensor based HAR. In this study, a Bidirectional Long Short-Term Memory (BLSTM) approach is explored for human activity recognition and classification for closely related activities on a body worn inertial sensor data that is provided by the UTD-MHAD dataset. The BLSTM model of this study could achieve an overall accuracy of 98.05% for 15 different activities and 90.87% for 27 different activities performed by 8 persons with 4 …

Infrastructure-Based Access Policy Enforcement Using Software-Defined Networks, Tyler Carden Culp

Theses and Dissertations

This thesis describes a method to enhance network security using software defined networks. Standard networks use perimeter-based defenses to block attackers from gaining access to internal systems. A key problem with standard networks is that once a malicious entity has gained access to the network, they are often able to freely move throughout the network and to attack internal systems with impunity. This problem can be mitigated by placing defenses such as firewalls between machines on the network, but this approach requires significant resources and constant maintenance. If the network infrastructure itself is leveraged as a defense by individualizing the …

Scale-Out Algorithm For Apache Storm In Saas Environment, Ravi Kiran Puttaswamy

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The main appeal of the Cloud is in its cost effective and flexible access to computing power. Apache Storm is a data processing framework used to process streaming data. In our work we explore the possibility of offering Apache Storm as a software service. Further, we take advantage of the cgroups feature in Storm to divide the computing power of worker machine into smaller units to be offered to users. We predict that the compute bounds placed on the cgroups could be used to approximate the state of the workflow. We discuss the limitations of the current schedulers in facilitating …

Reducing The Tail Latency Of A Distributed Nosql Database, Jun Wu

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The request latency is an important performance metric of a distributed database, such as the popular Apache Cassandra, because of its direct impact on the user experience. Specifically, the latency of a read or write request is defined as the total time interval from the instant when a user makes the request to the instant when the user receives the request, and it involves not only the actual read or write time at a specific database node, but also various types of latency introduced by the distributed mechanism of the database. Most of the current work focuses only on reducing …

Penetration Depth Variation In Atomic Layer Deposition On Multiwalled Carbon Nanotube Forests, David Alan Kane

Theses and Dissertations

Atomic Layer Deposition (ALD) of Al2O3 on tall multiwalled carbon nanotube forests shows concentration variation with the depth in the form of discrete steps. While ALD is capable of extremely conformal deposition in high aspect ratio structures, decreasing penetration depth has been observed over multiple thermal ALD cycles on 1.3 mm tall multiwalled carbon nanotube forests. SEM imaging with Energy Dispersive X-ray Spectroscopy elemental analysis shows steps of decreasing intensity corresponding to decreasing concentrations of Al2O3. A study of these steps suggests that they are produced by a combination of diffusion limited delivery of precursors with increasing precursor adsorption site …

Framing Climate Change As National Security: A Survey-Experiment On Climate Change Perceptions., Adam Lyons

Electronic Theses and Dissertations

The issues surrounding climate change continue to have polarizing effects on many Americans. In this thesis I offer a potential bridge to this divide with a comprehensive study on how issue frames can impact how individuals shape their opinions on environmental issues. Focusing on a national security frame, I ask the central research “Will the public perceive issues of climate change differently when they are framed as threats to national security?” For this thesis I design an original experimental survey to measure the impact security framing has on participants’ perception of both climate change and environmental policy options. The survey …

Cake Cam: Take Your Photo And Be In It Too, Candice Lynn Davis

Theses and Dissertations

In this thesis we explore a different kind of computer-mediated collaboration (CMC) in which two people collaborate to perform a single task in a time-sensitive asymmetric situation wherein one person is more invested in the outcome than the other is. Our conjecture is that interactive computing can quickly communicate intent, which may help mediate time-sensitive asymmetric tasks between two people. We explore this idea specifically in the context of asking a stranger to take one's picture at a tourist site. Consider a tourist handing a cell phone to a complete stranger and asking the stranger to take the tourist's picture …

Discovery Of Paleotsunami Deposits Along Eastern Sunda Arc: Potential For Megathrust Earthquakes In Bali, Hanif Ibadurrahman Sulaeman

Theses and Dissertations

Several laterally extensive candidate tsunami deposits are preserved along coastlines facing the eastern Java Trench, indicating it has experienced mega-thrust earthquakes in the past. We investigated 37 coastal sites in Bali, Lombok, Sumba and Timor islands, many of which preserve course sand and pebble layers that overlie sharp basal contacts with scour marks into the mud, fine upward in grain size, and have bimodal grain size distributions. Other unique features are the common occurrence of marine fossils and concentrations of heavy minerals. The occurrence of these high-energy deposits interlayered with clay-rich units indicates the coarse clastics are anomalous because they …

The Bird And The Fish: Motion Field-Based Frame Interpolation In The Context Of A Story, Thomas Sterling Ellsworth

Theses and Dissertations

Animating by hand can be a long and challenging process in part because of the necessity of drawing every frame by hand. 3D animation media minimize this problem with the use of automatically interpolated frames, but despite significant research no universally acceptable techniques have been demonstrated for 2 dimensional interpolation. In this paper we explore computer-assisted optimizations to the animation pipeline. Specifically, we utilize 3D motion fields to create more realistic in-between frames for sets of 2D ``key frames.'' We demonstrate our method by using it to create 2D special effects for a 30-second clip of an animated short film.

Improving The Quality Of Neural Machine Translation Using Terminology Injection, Duane K. Dougal

Theses and Dissertations

Most organizations use an increasing number of domain- or organization-specific words and phrases. A translation process, whether human or automated, must also be able to accurately and efficiently use these specific multilingual terminology collections. However, comparatively little has been done to explore the use of vetted terminology as an input to machine translation (MT) for improved results. In fact, no single established process currently exists to integrate terminology into MT as a general practice, and especially no established process for neural machine translation (NMT) exists to ensure that the translation of individual terms is consistent with an approved terminology collection. …

A Plausible Resolution To Hilbert’S Failed Attempt To Unify Gravitation & Electromagnetism, Florentin Smarandache, Victor Christianto, Robert Neil Boyd

Branch Mathematics and Statistics Faculty and Staff Publications

In this paper, we explore the reasons why Hilbert’s axiomatic program to unify gravitation theory and electromagnetism failed and outline a plausible resolution of this problem. The latter is based on Gödel’s incompleteness theorem and Newton’s aether stream model.

Usable Security And Privacy For Secure Messaging Applications, Elham Vaziripour

Theses and Dissertations

The threat of government and corporate surveillance around the world, as well as the publicity surrounding major cybersecurity attacks, have increased interest in secure and private end-to-end communications. In response to this demand, numerous secure messaging applications have been developed in recent years. These applications have been welcomed and publically used not just by political activists and journalists but by everyday users as well. Most of these popular secure messaging applications are usable because they hide many of the details of how encryption is provided. The strength of the security properties of these applications relies on the authentication ceremony, wherein …

Identification Of Receptor Ligands In Apo B100 Reveals Potential Functional Domains, Juan Guevara Jr., Jamie Romo Jr., Ernesto Hernandez, Natalia Valentinova Guevara

Physics and Astronomy Faculty Publications and Presentations

LDL, VLDL and other members of the low-density lipoparticles (LLPs) enter cells through a large family of receptors. The actual receptor ligand(s) in apolipoprotein B100, one of the main proteins of LLP, remain(s) unknown. The objective of this study was to identify true receptor ligand(s) in apo B100, a molecule of 4,563 residues. Apo B100 contains 33 analogues of Cardin-Weintraub Arginine/Lysine-based receptor ligand motifs and shares key Lysine motifs and sequence similarity with the LDL receptor-associated protein, RAP, MESD, and heat shock proteins. Eleven FITC-labeled synthetic peptides of 21 – 42 residues, with at least one ligand, were tested for …

A Normal Form For Words In The Temperley-Lieb Algebra And The Artin Braid Group On Three Strands, Jack Hartsell

Electronic Theses and Dissertations

The motivation for this thesis is the computer-assisted calculation of the Jones poly- nomial from braid words in the Artin braid group on three strands, denoted B3. The method used for calculation of the Jones polynomial is the original method that was created when the Jones polynomial was first discovered by Vaughan Jones in 1984. This method utilizes the Temperley-Lieb algebra, and in our case the Temperley-Lieb Algebra on three strands, denoted A3, thus generalizations about A3 that assist with the process of calculation are pursued.

Measurement Of The Cross Section Of Top Quark Pairs In Association With A Photon In Lepton+Jets Events At √S =13 Tev, Titas Roy

Theses and Dissertations

The Large Hadron Collider (LHC) at CERN is the world’s biggest and most powerful accelerator. Colliding protons at the highest centre of mass energy yet, data are being collected by the detectors at an unprecendented rate, making it possible to search for new particles as well as make precise measurements. With the increase in statistics and energy of the data collected, it is now possible to make precise measurements with events that have a relatively low probability of occuring. In this thesis, the production cross section of a top quark pair plus a radiated photon, tt¯+γ, is measured in proton-proton …

Lighting A Path To Biofouling Prevention: Investigating The Effect Of Ultraviolet Light On Biofouling, Cierra Rose Braga

Theses and Dissertations

Ultraviolet light (UV) is considered an environmentally friendly antifouling method because it disrupts biological growth without the release of harmful toxins. The application of UV for the marine shipping industry is a relatively new concept, and there still remains many unknowns before it can be fully implemented. This research investigated several aspects of applying UVC for the control of fouling. A novel method of using embedded UV LEDs within a silicone coating for fouling control proved to be very effective at preventing fouling. The research also investigated the effect of using an external UVC light treatment to control fouling on …

Wind-Driven Setup, Roughness, And Drag Over A Depth- And Fetch-Limited Coastal Estuary, Jeffrey Allen Colvin

Theses and Dissertations

The understanding of the momentum exchange across the air-sea interface is critical to accurately model and forecast wind speed, wave properties, setup, and circulation. This is particularly true for a restricted estuary, such as the Indian River Lagoon (IRL), where wind is the driving force behind water movement due to the limited number of direct ocean connections. Current atmospheric and hydrodynamic models use formulations of the drag coefficient and/or roughness length based on open ocean studies in which wave properties are not considered. This dissertation develops two parameterizations – one for wind setup and one for the drag coefficient – …

Mitigating Memory Randomization Weaknesses Via Moving Target Defense, Sultan M. Aldossary

Theses and Dissertations

In this dissertation, we address the problem of weakness of address space layout randomization. The address space randomization technique was proposed to make determining the address of a shared library more difficult since each instance of the program is loaded into a random base address. However, when address space randomization layout (ASLR) is implemented on a 32-bit system, an attacker can use a brute force attack to guess the address of the shared library.The main goal of the research described in this dissertation is to study the use of a dispatching algorithm and multiple back-end servers as a moving target …

Collection And Analysis Of Digital Forensic Data From Devices In The Internet Of Things, Raed Alharbi

Theses and Dissertations

Despite the abundance articles that have been written about the Internet of Things (IoT), little attention has been given to how digital forensics approaches can be utilized to direct advanced investigations in IoT-based frameworks. As of yet, IoT has not completely adjusted to digital forensic strategies given the fact that current digital forensic tools and functions are not ready to tackle the complexity of IoT frameworks for the purpose of collecting, analyzing, and testing potential evidence from IoT environments that might be utilized as permissible evidence in a court. Hence, the issue addressed is that; currently, there is no accepted …

Automatic Anuran Species Recognition Via Vocalization, Katrina Smart

Theses and Dissertations

Species recognition of Anuran vocalization is an important problem in ecology, which is costly and time consuming. Recently, biologists have turned to machine learning techniques to automate the process. An automated process can be made even more helpful if available to use on a smart phone, which can be used in the field by scientists and citizen scientists alike. The problem of automatic frog call recognition is challenging because features need to be chosen that can properly separate different species’ very similar sounding vocalizations. This thesis offers the following contributions: 1. Three novel feature extraction methods, which do not require …

Development Of Portable Embedded Face Recognition Prototype, Prasad S. Garapati

Theses and Dissertations

In this thesis, we focus on developing a prototype for collecting multispectral images for face recognition. With such application, the quality of the image is an important factor that affects the accuracy of the recognition. However, the sensory data either provided in RGB channel or in NIR channel. The images acquired in RGB and NIR were never aligned in previous developments during my literature study. Thus, I propose an embedded system prototype capable of acquiring NIR and RGB images providing multispectral information which is computation efficient for real time face recognition applications.

A New Generation Of Functional Polyisobutylenes For Advanced Applications, Corey M. Parada

Dissertations

Polyisobutylene (PIB) is a fully saturated, aliphatic polymer of high commercial importance due to its superior gas barrier properties and high chemical/oxidative stability. One commercial end-use for PIB is in insulated glass windows (IGU), where it acts as a gas/moisture barrier and sealant. Under certain adverse conditions, catastrophic failure of the PIB sealant may result in aesthetic and functional failure of the IGU, which necessitates replacement of the unit. Thus, there exists a need to improve current generations of thermoplastic PIB sealants to be able to withstand the harsh environments found in current real-world applications.

In the first project, we …

Polyhedral Oligomeric Silsesquioxane-Phosphate Glass Matrix Nanocomposites With Additional Chapter On Phosphate Glass-Poly(Ethylene Terephthalate) Matrix Composites, Kyoungtae Kim

Dissertations

Preparation and characterization of tin fluorophosphate glass (Pglass) matrix nanocomposites incorporating polyhedral oligomeric silsesquioxane (POSS) were investigated on the structural, thermal, morphological, mechanical, and rheological properties. Various processes including synthesis, extrusion and sintering processes were applied to prepare the nanocomposite samples. Another application of POSS with hydrophobic functional groups on the well-structured nanoscale silicate cage with three silanol groups was investigated to present the feasibility to use POSS molecule as a coating material on the surface of the hydrophilic inorganic glass. In addition, Poly(ethylene terephthalate) polymer matrix composites incorporating Pglass was studied to present the benefits of the Pglass with …

Predicted Deepwater Bathymetry From Satellite Altimetry: Non-Fourier Transform Alternatives, Maxsimo Salazar

Dissertations

Robert Parker (1972) demonstrated the effectiveness of Fourier Transforms (FT) to compute gravitational potential anomalies caused by uneven, non-uniform layers of material. This important calculation relates the gravitational potential anomaly to sea-floor topography. As outlined by Sandwell and Smith (1997), a six-step procedure, utilizing the FT, then demonstrated how satellite altimetry measurements of marine geoid height are inverted into seafloor topography. However, FTs are not local in space and produce Gibb’s phenomenon around discontinuities. Seafloor features exhibit spatial locality and features such as seamounts and ridges often have sharp inclines. Initial tests compared the windowed-FT to wavelets in reconstruction of …

Effect Of Network Structure On Free Volume And Gas Transport Properties Of Thiol-Ene And Epoxy-Amine Networks, Ramesh Krishnan Ramakrishnan

Dissertations

Network parameters such as cross-link density or intermolecular interactions were used as effective parameters to control polymer chain packing or free volume and thereby control the mass transport properties of networks.

A series of polyethylene glycol (PEG) based thiol-ene elastomeric networks having a broad range of molecular weight between cross-links, M_c, (inverse of cross-link density) was prepared. The specific volume of the networks was studied as a function of temperature and pressure using high pressure dilatometry, and the obtained PVT data was fit using Simha-Somcynsky equation of state (S-S eos) analysis. Fractional free volume, quantified through S-S eos …

Structure-Property Relationships Of Solvent-Cast Organic Thin Films For Model Corrosion Control Coatings, Austin Maples

Dissertations

This dissertation aimed at answering several fundamental questions, specifically: 1) What is the effect of scribing parameters on visual corrosion assessments? 2) What role do molecular weight, residual solvent, and polymer composition have on morphological changes in organic thin films? 3) What alternatives to visual assessments could be implemented to better understand and detect corrosion before macroscopic failure? This research definitively proved that scribes can vary in type and dimensions, and yet are still comparable if the scribe dimensions before exposure are known. Furthermore, it demonstrated that solvents common to a variety of coating industries could become trapped within the …