Physical Sciences and Mathematics Commons^™

Transfer Learning: Bridging The Gap Between Deep Learning And Domain-Specific Text Mining, Chaoran Cheng

Dissertations

Inspired by the success of deep learning techniques in Natural Language Processing (NLP), this dissertation tackles the domain-specific text mining problems for which the generic deep learning approaches would fail. More specifically, the domain-specific problems are: (1) success prediction in crowdfunding, (2) variants identification in biomedical literature, and (3) text data augmentation for domains with low-resources.

In the first part, transfer learning in a multimodal perspective is utilized to facilitate solving the project success prediction on the crowdfunding application. Even though the information in a project profile can be of different modalities such as text, images, and metadata, most existing …

Stress-State And Injection-Rate Dependent Damage Processes During The Hydraulic Fracturing Of Granite, Gayani Sasendrika Gunarathna

Dissertations

Hydraulic fracturing is a well-stimulation technique that is employed in field applications, such as enhanced geothermal systems (EGS) and shale oil/gas extraction. This research experimentally investigates the effect of the state of stress and injection rate on the hydraulic fracturing processes. In addition, a displacement discontinuity method (DDM) code, FROCK, is used to model the crack initiation and propagation in a granite specimen under hydraulic fracturing conditions. In order to conduct the experimental work, a test setup capable of applying a triaxial state of stress and water pressure inside pre-fabricated flaws cut in prismatic granite specimens is developed. Additionally, the …

Modeling Single Microtubules As A Colloidal System To Measure The Harmonic Interactions Between Tubulin Dimers In Bovine Brain Derived Versus Cancer Cell Derived Microtubules, Arooj Aslam

Dissertations

The local properties of tubulin dimers dictate the properties of the larger microtubule assembly. In order to elucidate this connection, tubulin-tubulin interactions are be modeled as harmonic interactions to map the stiffness matrix along the length of the microtubule. The strength of the interactions are measured by imaging and tracking the movement of segments along the microtubule over time, and then performing a fourier transform to extract the natural vibrational frequencies. Using this method the first ever reported experimental phonon spectrum of the microtubule is reported. This method can also be applied to other biological materials, and opens new doors …

Subspace Portfolios: Design And Performance Comparison, Anqi Xiong

Dissertations

Data processing and engineering techniques enable people to observe and better understand the natural and human-made systems and processes that generate huge amounts of various data types. Data engineers collect data created in almost all fields and formats, such as images, audio, and text streams, biological and financial signals, sensing and many others. They develop and implement state-of-the art machine learning (ML) and artificial intelligence (AI) algorithms using big data to infer valuable information with social and economic value. Furthermore, ML/AI methodologies lead to automate many decision making processes with real-time applications serving people and businesses. As an example, mathematical …

Efficient Hardware Implementations Of Bio-Inspired Networks, Anakha Vasanthakumaribabu

Dissertations

The human brain, with its massive computational capability and power efficiency in small form factor, continues to inspire the ultimate goal of building machines that can perform tasks without being explicitly programmed. In an effort to mimic the natural information processing paradigms observed in the brain, several neural network generations have been proposed over the years. Among the neural networks inspired by biology, second-generation Artificial or Deep Neural Networks (ANNs/DNNs) use memoryless neuron models and have shown unprecedented success surpassing humans in a wide variety of tasks. Unlike ANNs, third-generation Spiking Neural Networks (SNNs) closely mimic biological neurons by operating …

Nanocarbon Modification Of Membranes For Enhanced Water Desalination And Water Treatment, Worawit Intrchom

Dissertations

Water scarcity is foreseen to be one of the great global issues in the coming decades. The challenges are not only in providing water supply to cope with the growing public demand, but recovering clean water to natural resources. Clean water supply, from brackish and seawater is attractive. Membrane distillation (MD) is an emerging thermal membrane-based process that has been used for desalination and other pollutant separations from water. MD can be operated at low temperature, so low-grade energy sources are a good alternative heat source for MD. High salt rejection and low membrane fouling also make MD interesting for …

Data Assimilation For Conductance-Based Neuronal Models, Matthew Moye

Dissertations

This dissertation illustrates the use of data assimilation algorithms to estimate unobserved variables and unknown parameters of conductance-based neuronal models. Modern data assimilation (DA) techniques are widely used in climate science and weather prediction, but have only recently begun to be applied in neuroscience. The two main classes of DA techniques are sequential methods and variational methods. Throughout this work, twin experiments, where the data is synthetically generated from output of the model, are used to validate use of these techniques for conductance-based models observing only the voltage trace. In Chapter 1, these techniques are described in detail and the …

Numerical And Analytical Methods To Predict Behavior Of Reinforced Ductile Concrete Composites, Mandeep Pokhrel

Dissertations

Structural components constructed with ductile concrete composites, such as high-performance fiber-reinforced cementitious composites (HPFRCC), are known to perform exceptionally well under extreme mechanical and environmental loading conditions compared to traditional concrete. HPFRCC flexural components exhibit enhanced performance in terms of displacement ductility, energy dissipation capacity, and damage tolerance capacity. However, recent research suggests that the flexural behavior of reinforced HPFRCCs in terms of crack progression, reinforcement plasticity, and failure mechanism is significantly different than conventional reinforced concrete. Specifically, the failure mode of flexural members is found to be predominantly through the fracture of longitudinal reinforcement rather than compression crushing of …

Coding Against Stragglers In Distributed Computation Scenarios, Malihe Aliasgari

Dissertations

Data and analytics capabilities have made a leap forward in recent years. The volume of available data has grown exponentially. The huge amount of data needs to be transferred and stored with extremely high reliability. The concept of "coded computing", or a distributed computing paradigm that utilizes coding theory to smartly inject and leverage data/computation redundancy into distributed computing systems, mitigates the fundamental performance bottlenecks for running large-scale data analytics.

In this dissertation, a distributed computing framework, first for input files distributedly stored on the uplink of a cloud radio access network architecture, is studied. It focuses on that decoding …

Calculating Elastic Properties Of Confined Simple Fluids, Christopher D. Dobrzanski

Dissertations

Confinement in nanoporous materials is known to affect many properties of the fluids confined within their pores. The elastic properties are no exception. This dissertation begins with an overview of the relevant literature on ways of obtaining elastic properties of confined fluids. It outlines some fundamental gaps in our understanding. The chapters following address some of these gaps in understanding elastic properties of the confined fluid, in particular, how the shape of the confining pore matters, how supercriticality effects the properties, how an equation of state designed for confined fluids can be used to calculate elastic properties, and if an …

Construct Local Quasi-Interpolation Operators Using Linear B-Splines, Abeer Alzahrani

Dissertations

The data interpolating problem is a fundamental problem in data analysis, and B-splines are frequently used as the basis functions for data interpolation. In the real-world applications, the real-time processing is very important. To achieve that, we cannot use any matrix inversion for large amount of data, and we also need to avoid using any global operator. To solve this problem, we develop a new method based on a local quasi-interpolation operator. To construct the local quasi-interpolation operator, we need to factorize the Shoenberg-Whitney matri- ces for the given data samples. Furthermore, our local quasi-interpolation operator should correspond to a …

An Atomistic Study Of The Effects On Mechanical Properties And Bonding Interactions Of Carbon Nanofillers In Nylon 6 Nanocomposites, Michael Roth

Dissertations

Polymers have potential for a wide range of applications. The effectiveness of polymers can be further enhanced through the addition of nanofillers that improve thermal, mechanical, and electrical properties of the polymer. Carbon based nanofillers such as carbon nanotube (CNT), graphene, and carbon nanofibre (CNF) are of particular interest due to their high properties and high aspect ratios. However, limited understanding of the governing interactions of these nanofillers with polymers limits the effectiveness of the final nanocomposite.

The first facet of this dissertation focuses on determining the dominating interactions between pristine CNT and graphene with nylon 6 monomer and the …

Effect Of Selfsame Microparticles On Epoxide Amine Network Formation And Matrix Mechanics, Travis Palmer

Dissertations

Epoxide amine matrices are widely utilized in aerospace carbon fiber reinforced polymer (CFRP) composites having engendered significant reductions in weight and fuel consumption. This dissertation focuses on the effect of constrained space during network formation on the matrix mechanics of these highly complex composite systems. Precipitation polymerization conditions are developed to prepare epoxide amine microparticles (EMs) based on tetraglycidyl-4,4’-methylenedianiline (TGDDM) and isophorone diamine (IPDA). Surface functionality of EMs is tuned via control of epoxide to reactive amine hydrogen ratio, where unreactive, amine- and epoxide-functional EMs are prepared. We demonstrate that EMs are polydisperse, but can be filtered, yielding low dispersity …

A Dynamic F5 Algorithm, Candice Mitchell

Dissertations

Gröbner bases are a “nice” representation for nonlinear systems of polynomials, where by “nice” we mean they have good computation properties. They have many useful applications, including decidability (whether the system has a solution or not), ideal membership (whether a given polynomial is in the system or not), and cryptography. Traditional Gröbner basis algorithms require as input an ideal and an admissible term ordering. They then determine a Gröbner basis with respect to the given ordering. Some term orderings lead to a smaller basis, but finding them traditionally requires testing many orderings and hoping for better results. A dynamic algorithm …

Huisgen 1,3-Dipolar Azide-Alkyne Cycloaddition “Click” Reaction In Polymer Synthesis And Curing, Jie Wu

Dissertations

This dissertation’s key focus is on utilizing Huisgen 1,3-dipolar azide-alkyne cycloaddition (AAC) reaction in copolymer synthesis and modification, including thermoplastic block copolymer and commercially available two-component polyurethane system. It can be divided into two major projects, introduced as follows.

The first project involves the development of a modular synthetic approach toward polyisobutylene (PIB)-based triphasic pentablock thermoplastic elastomer with enhanced moisture permeability. This terpolymer consists of a poly(styrene-b-isobutylene-b-styrene) (SIBS) core and appended hydrophilic polymer blocks (HBs). The SIBS core was synthesized via living cationic polymerization (LCP) of isobutylene followed by sequential addition of styrene. AAC was utilized …

Geochemical Tracers Of Arctic Ocean Processes: A Study Of Gallium, Barium, And Vanadium, Laura M. Whitmore

Dissertations

The Arctic Ocean is linked to the global oceans and climate through its connectivity with the North Atlantic Ocean and the regional thermohaline deep water formation sites. It’s also a region undergoing rapid environmental change. To inform the community of potential changes in geochemical and biogeochemical cycles, this dissertation addresses three dissolved geochemical tracers (gallium, barium, and vanadium) as indicators of Arctic Ocean processes. Gallium is tested as a replacement for nutrient-type tracers in an effort to deconvolve Pacific and Atlantic derived waters in the Arctic Ocean basins. These water masses carry different heat and salt content and can influence …

Cross Language Information Transfer Between Modern Standard Arabic And Its Dialects – A Framework For Automatic Speech Recognition System Language Model, Tiba Zaki Abdulhameed

Dissertations

Significant advances have been made with Modern Standard Arabic (MSA) Automatic Speech Recognition (ASR) applications. Yet, dialectal conversation ASR is still trailing behind due to limited language resources. As is the case in most cultures, the formal Modern Standard Arabic language is not used in daily life. Instead, varieties of regional dialects are spoken, which creates a dire need to address dialect ASR systems. Processing MSA language naturally poses considerable challenges that are passed on to the processing of its derived dialects. In dialects, many words have gradually morphed from MSA pronunciations and at many times have different usages. Also, …

Activated Carbon Nanofibers From Renewable (Lignin) And Waste Resources (Recycled Pet) And Their Adsorption Capacity Of Refractory Sulfur Compounds From Fossil Fuels, Efstratios Svinterikos

Dissertations

Dementia is a condition in which higher mental functions are disrupted. It currently affects an estimated 57 million people throughout the world. Dementia diagnosis is difficult since neither anatomical indicator nor functional testing are currently sufficiently sensitive or specific. There remains a long list of outstanding issues that must be addressed. First, multimodal diagnosis has yet to be introduced into the early stages of dementia screening. Second, there is no accurate instrument for predicting the progression of pre-dementia. Third, non-invasive testing cannot be used to provide differential diagnoses. By creating ML models of normal and accelerated brain aging, we intend …

Assessment Of The Impact Of Climate Change On Land Use In The Emirate Of Abu Dhabi - An Environmental And Socio-Economic Perspective, Latifa Saeed Al Blooshi

Dissertations

This dissertation focuses on the impact of climate change on land use in the Emirate of Abu Dhabi – UAE. Climate change is a significant challenge resulting from natural and anthropogenic causes. Land use can stimulate changes in communities under climate change. The main objective of this dissertation is to assess the impact of climate change from an environmental and socio-economic perspective. In 2001, coastal sabkhas, mixed class and urbanized areas experienced an increase in temperature by (0.67, 1.14 and 1.16°C) respectively. In cities, urban areas are warmer than neighbouring rural areas. Unexpectedly, urbanization in desert areas in UAE led …

Computer Simulations Of Muscle Driven Undulatory Locomotion, Ye Luo

Dissertations

A novel muscle driven method is developed to mimic contracting and expanding of muscles, in a fish-like swimming body, which cause its body flapping in the transversal direction and create thrust force to push its body to cruise in the longitudinal direction. The muscle deformation is realized by using the RATTLE constraint algorithm. The turbulent fluids are treated by a multi-relaxation time lattice Boltzmann method with a large eddy simulation. The fish body is dealt with a lattice spring model and interactions between fluids and solid structures are handled by a direct-forcing immersed boundary method. Validations are conducted by comparing …

Extremal Problems On Induced Graph Colorings, James Hallas

Dissertations

Graph coloring is one of the most popular areas of graph theory, no doubt due to its many fascinating problems and applications to modern society, as well as the sheer mathematical beauty of the subject. As far back as 1880, in an attempt to solve the famous Four Color Problem, there have been numerous examples of certain types of graph colorings that have generated other graph colorings of interest. These types of colorings only gained momentum a century later, however, when in the 1980s, edge colorings were studied that led to vertex colorings of various types, led by the introduction …

High Performance And Machine Learning Algorithms For Brain Fmri Data, Taban Eslami

Dissertations

Brain disorders are very difficult to diagnose for reasons such as overlapping nature of symptoms, individual differences in brain structure, lack of medical tests and unknown causes of some disorders. The current psychiatric diagnostic process is based on behavioral observation and may be prone to misdiagnosis.

Noninvasive brain imaging technologies such as Magnetic Resonance Imaging (MRI) and functional Magnetic Resonance Imaging (fMRI) make the process of understanding the structure and function of the brain easier. Quantitative analysis of brain imaging data using machine learning and data mining techniques can be advantageous not only to increase the accuracy of brain disorder …

New Catalytic Reactions In Carbohydrate Chemistry, Scott Geringer

Dissertations

Carbohydrates or sugars are some of the most diverse and abundant biological molecules. They are involved in a multitude of processes in the body such as fertilization, cell-cell communication, and cancer metathesis. Because of these vital functions, the study of sugars is rapidly growing field. The field however is limited due to the complex nature of sugars which results in difficulties in obtaining large quantities for study.

Protecting group manipulation is a large emphasis area in carbohydrate chemistry due to the need to selectively protect different functional groups of each molecule during synthesis. Catalytic and selective cleavage of protecting groups …

High-Order Adaptive Synchrosqueezing Transform, Jawaher Alzahrani

Dissertations

The prevalence of the separation of multicomponent non-stationary signals across many elds of research makes this concept an important subject of study. The synchrosqueezing transform (SST) is a particular type of reassignment method. It aims to separate and recover the components of a multicomponent non-stationary signal. The short time Fourier transform (STFT)-based SST (FSST) and the continuous wavelet transform (CWT)based SST (WSST) have been used in engineering and medical data analysis applications. The current study introduces the dierent versions of FSST and WSST to estimate instantaneous frequency (IF) and to recover components. It has a good concentration and reconstruction for …

Noise Reduction Of Eeg Signals Using Autoencoders Built Upon Gru Based Rnn Layers, Esra Aynali

Dissertations

Understanding the cognitive and functional behaviour of the brain by its electrical activity is an important area of research. Electroencephalography (EEG) is a method that measures and record electrical activities of the brain from the scalp. It has been used for pathology analysis, emotion recognition, clinical and cognitive research, diagnosing various neurological and psychiatric disorders and for other applications. Since the EEG signals are sensitive to activities other than the brain ones, such as eye blinking, eye movement, head movement, etc., it is not possible to record EEG signals without any noise. Thus, it is very important to use an …

Metal Ions Impact On Shewanella Oneidensis Mr-1 Adhesion To Ito Electrode And The Enhancement Of Current Output, Aisha Awad Alshahrani

Dissertations

The goal of this study is to enhance the efficiency of bacterial extracellular electron transfer (EET) in Shewanella oneidensis MR-1 by enhancing adhesion to the electrode's surface. Our results clearly show a major difference in the attachment and behavior of Shewanella oneidensis MR-1 for Ca2+, Pb2+, Cd2+, and Mg2+, compared to the control. the final microbial coverage, as measured by confocal microscopy and cathodic peak charge in cyclic voltammetry (Qpc), increases with increasing metal ion concentrations. We found the cells attached to the electrode increased more with the addition of metal ion concentrations in the following order of metals: Ca2+ …

Optimization Of Microbial Fuel Cells, Norberto M. Gonzalez

Dissertations

Optimization of microbial fuel cells is investigated by utilizing Shewanella oneidensis as a model microorganism. the microbe's ability to grow and use glucose as a carbon source is explored under varying oxygen environments through an offline PMP derivatization method and HPLC analysis. Shewanella growth and glucose utilization is enhanced under aerobic environments; however, under microaerobic environments the addition of ferric iron results in a faster exponential growth initialization. a flavin mononucleotide modified indium tin oxide electrode is prepared and characterized for its usefulness in microbial fuel cells by controlled potential electrolysis, cyclic voltammetry, and electrochemical impedance spectroscopy studies. the electrode …

Electron Transfer Of Shewanella Oneidensis Mr-1 At Clay-Modified Ito Electrode, Reem Faez Alshehri

Dissertations

Various strategies have been established to enhance the extracellular electron transfer and energy output capability of microbial fuel cells, with the majority being aimed at anode modification. The anode has a significant impact on the electricity generation performance of MFCs because it is in direct contact with the microorganisms. The materials of the anode should be favorable for the bacterial cell and capable to facilitate the electron transfer. Developing of an electrode using low-cost and effective materials assists to enhance the bacterial cell attachment and extracellular electron transfer. This provides a significant improvement in MFC performance. In this study, clay …

Brain Disease Detection From Eegs: Comparing Spiking And Recurrent Neural Networks For Non-Stationary Time Series Classification, Hristo Stoev

Dissertations

Modeling non-stationary time series data is a difficult problem area in AI, due to the fact that the statistical properties of the data change as the time series progresses. This complicates the classification of non-stationary time series, which is a method used in the detection of brain diseases from EEGs. Various techniques have been developed in the field of deep learning for tackling this problem, with recurrent neural networks (RNN) approaches utilising Long short-term memory (LSTM) architectures achieving a high degree of success. This study implements a new, spiking neural network-based approach to time series classification for the purpose of …

Image Instance Segmentation: Using The Cirsy System To Identify Small Objects In Low Resolution Images, Orghomisan William Omatsone

Dissertations

The CIRSY system (or Chick Instance Recognition System) is am image processing system developed as part of this research to detect images of chicks in highly-populated images that uses the leading algorithm in instance segmentation tasks, called the Mask R-CNN. It extends on the Faster R-CNN framework used in object detection tasks, and this extension adds a branch to predict the mask of an object along with the bounding box prediction. Mask R-CNN has proven to be effective ininstance segmentation and object de-tection tasks after outperforming all existing models on evaluation of the Microsoft Common Objects in Context (MS COCO) …