Physical Sciences and Mathematics Commons^™

Towards A Robust Defense: A Multifaceted Approach To The Detection And Mitigation Of Neural Backdoor Attacks Through Feature Space Exploration And Analysis, Liuwan Zhu

Electrical & Computer Engineering Theses & Dissertations

From voice assistants to self-driving vehicles, machine learning(ML), especially deep learning, revolutionizes the way we work and live, through the wide adoption in a broad range of applications. Unfortunately, this widespread use makes deep learning-based systems a desirable target for cyberattacks, such as generating adversarial examples to fool a deep learning system to make wrong decisions. In particular, many recent studies have revealed that attackers can corrupt the training of a deep learning model, e.g., through data poisoning, or distribute a deep learning model they created with “backdoors” planted, e.g., distributed as part of a software library, so that the …

Wearable Sensor Gait Analysis For Fall Detection Using Deep Learning Methods, Haben Girmay Yhdego

Electrical & Computer Engineering Theses & Dissertations

World Health Organization (WHO) data show that around 684,000 people die from falls yearly, making it the second-highest mortality rate after traffic accidents [1]. Early detection of falls, followed by pneumatic protection, is one of the most effective means of ensuring the safety of the elderly. In light of the recent widespread adoption of wearable sensors, it has become increasingly critical that fall detection models are developed that can effectively process large and sequential sensor signal data. Several researchers have recently developed fall detection algorithms based on wearable sensor data. However, real-time fall detection remains challenging because of the wide …

Cyber Resilience Analytics For Cyber-Physical Systems, Md Ariful Haque

Electrical & Computer Engineering Theses & Dissertations

Cyber-physical systems (CPSs) are complex systems that evolve from the integrations of components dealing with physical processes and real-time computations, along with networking. CPSs often incorporate approaches merging from different scientific fields such as embedded systems, control systems, operational technology, information technology systems (ITS), and cybernetics. Today critical infrastructures (CIs) (e.g., energy systems, electric grids, etc.) and other CPSs (e.g., manufacturing industries, autonomous transportation systems, etc.) are experiencing challenges in dealing with cyberattacks. Major cybersecurity concerns are rising around CPSs because of their ever-growing use of information technology based automation. Often the security concerns are limited to probability-based possible attack …

Applied Deep Learning: Case Studies In Computer Vision And Natural Language Processing, Md Reshad Ul Hoque

Electrical & Computer Engineering Theses & Dissertations

Deep learning has proved to be successful for many computer vision and natural language processing applications. In this dissertation, three studies have been conducted to show the efficacy of deep learning models for computer vision and natural language processing. In the first study, an efficient deep learning model was proposed for seagrass scar detection in multispectral images which produced robust, accurate scars mappings. In the second study, an arithmetic deep learning model was developed to fuse multi-spectral images collected at different times with different resolutions to generate high-resolution images for downstream tasks including change detection, object detection, and land cover …

Emotion Detection Using An Ensemble Model Trained With Physiological Signals And Inferred Arousal-Valence States, Matthew Nathanael Gray

Electrical & Computer Engineering Theses & Dissertations

Affective computing is an exciting and transformative field that is gaining in popularity among psychologists, statisticians, and computer scientists. The ability of a machine to infer human emotion and mood, i.e. affective states, has the potential to greatly improve human-machine interaction in our increasingly digital world. In this work, an ensemble model methodology for detecting human emotions across multiple subjects is outlined. The Continuously Annotated Signals of Emotion (CASE) dataset, which is a dataset of physiological signals labeled with discrete emotions from video stimuli as well as subject-reported continuous emotions, arousal and valence, from the circumplex model, is used for …

Development Of High Quantum Efficiency Strained Superlattice Spin Polarized Photocathodes Via Metal Organic Chemical Vapor Deposition, Benjamin Belfore

Electrical & Computer Engineering Theses & Dissertations

Spin polarized photocathodes are necessary to examine parity violations and other fundamental phenomena in the field of high energy physics. To create these devices, expensive and complicated growth processes are necessary. While integral to accelerator physics, spin polarized electrons could have other exciting applications in materials science and other fields of physics. In order to explore these other applications feasibly, the relative supply of spin polarized photocathodes with a high rate of both polarization and photoemission needs to be increased. One such way to increase this supply is to develop the means to grow them faster and at a larger …

Chen-Fliess Series For Linear Distributed Systems, Natalie T. Pham

Electrical & Computer Engineering Theses & Dissertations

Distributed systems like fluid flow and heat transfer are modeled by partial differential equations (PDEs). In control theory, distributed systems are generally reformulated in terms of a linear state space realization, where the state space is an infinite dimensional Banach space or Hilbert space. In the finite dimension case, the input-output map can always be written in terms of a Chen-Fliess functional series, that is, a weighted sum of iterated integrals of the components of the input function. The Chen-Fliess functional series has been used to describe interconnected nonlinear systems, to solve system inversion and tracking problems, and to design …

Machine Learning Classification Of Digitally Modulated Signals, James A. Latshaw

Electrical & Computer Engineering Theses & Dissertations

Automatic classification of digitally modulated signals is a challenging problem that has traditionally been approached using signal processing tools such as log-likelihood algorithms for signal classification or cyclostationary signal analysis. These approaches are computationally intensive and cumbersome in general, and in recent years alternative approaches that use machine learning have been presented in the literature for automatic classification of digitally modulated signals. This thesis studies deep learning approaches for classifying digitally modulated signals that use deep artificial neural networks in conjunction with the canonical representation of digitally modulated signals in terms of in-phase and quadrature components. Specifically, capsule networks are …

Joint Linear And Nonlinear Computation With Data Encryption For Efficient Privacy-Preserving Deep Learning, Qiao Zhang

Electrical & Computer Engineering Theses & Dissertations

Deep Learning (DL) has shown unrivalled performance in many applications such as image classification, speech recognition, anomalous detection, and business analytics. While end users and enterprises own enormous data, DL talents and computing power are mostly gathered in technology giants having cloud servers. Thus, data owners, i.e., the clients, are motivated to outsource their data, along with computationally-intensive tasks, to the server in order to leverage the server’s abundant computation resources and DL talents for developing cost-effective DL solutions. However, trust is required between the server and the client to finish the computation tasks (e.g., conducting inference for the newly-input …

Deep Learning Approaches For Seagrass Detection In Multispectral Imagery, Kazi Aminul Islam

Electrical & Computer Engineering Theses & Dissertations

Seagrass forms the basis for critically important marine ecosystems. Seagrass is an important factor to balance marine ecological systems, and it is of great interest to monitor its distribution in different parts of the world. Remote sensing imagery is considered as an effective data modality based on which seagrass monitoring and quantification can be performed remotely. Traditionally, researchers utilized multispectral satellite images to map seagrass manually. Automatic machine learning techniques, especially deep learning algorithms, recently achieved state-of-the-art performances in many computer vision applications. This dissertation presents a set of deep learning models for seagrass detection in multispectral satellite images. It …

Wiener-Fliess Composition Of Formal Power Series: Additive Static Feedback And Shuffle Rational Series, Subbarao Venkatesh Guggilam

Electrical & Computer Engineering Theses & Dissertations

The problem statement for this dissertation is two-fold. The first problem considered is when does a Chen-Fliess series in an additive static feedback connection with a formal static map yield a closed-loop system with a Chen-Fliess series expansion? This work proves that such a closed-loop system always has a Chen-Fliess series representation. Furthermore, an algorithm based on the Hopf algebras for the shuffle group and the dynamic output feedback group is designed to compute the generating series of the closed-loop system. It is proved that the additive static feedback connection preserves local convergence and relative degree, but a counterexample shows …

Electrostatic Design And Characterization Of A 200 Kev Photogun And Wien Spin Rotator, Gabriel G. Palacios Serrano

Electrical & Computer Engineering Theses & Dissertations

High-energy nuclear physics experiments at the Jefferson Lab Continuous Electron Beam Accelerator Facility (CEBAF) require high spin-polarization electron beams produced from strained super-lattice GaAs photocathodes activated to negative electron affinity in a high voltage photogun operating at 130 kV dc. A pair of Wien filter spin rotators in the injector provides precise control of the electron beam polarization at the end station target. An upgrade of the CEBAF injector to better support the upcoming Moller experiment requires increasing the electron beam energy to 200 keV, resulting in better transmission through injector apertures and improved photocathode lifetime. In addition, the energy …

Commissioning & Characterization Of Magnetized Gridded Thermionic Electron Source, Mark Stefani

Electrical & Computer Engineering Theses & Dissertations

Collaborative efforts to design and fabricate a magnetized gridded thermionic electron source have been conducted between Xelera and Jefferson Lab. Commissioning and characterization of an electron gun fabricated by Xelera was performed to benchmark the viability of future electron source designs and capabilities. The work involved simulation, installation, trouble-shooting, modifications of the design, commissioning, characterization, and magnetization of the electron beam produced. A specially designed cavity as well as novel diagnostic tools and methods were developed, implemented, and experimentally tested. Finally, the gun was used to demonstrate a previously unachieved current of magnetized electron beam from a gridded thermionic source …

Secure Mobile Computing By Using Convolutional And Capsule Deep Neural Networks, Rui Ning

Electrical & Computer Engineering Theses & Dissertations

Mobile devices are becoming smarter to satisfy modern user's increasing needs better, which is achieved by equipping divers of sensors and integrating the most cutting-edge Deep Learning (DL) techniques. As a sophisticated system, it is often vulnerable to multiple attacks (side-channel attacks, neural backdoor, etc.). This dissertation proposes solutions to maintain the cyber-hygiene of the DL-Based smartphone system by exploring possible vulnerabilities and developing countermeasures.

First, I actively explore possible vulnerabilities on the DL-Based smartphone system to develop proactive defense mechanisms. I discover a new side-channel attack on smartphones using the unrestricted magnetic sensor data. I demonstrate that attackers can …

Deep Cellular Recurrent Neural Architecture For Efficient Multidimensional Time-Series Data Processing, Lasitha S. Vidyaratne

Electrical & Computer Engineering Theses & Dissertations

Efficient processing of time series data is a fundamental yet challenging problem in pattern recognition. Though recent developments in machine learning and deep learning have enabled remarkable improvements in processing large scale datasets in many application domains, most are designed and regulated to handle inputs that are static in time. Many real-world data, such as in biomedical, surveillance and security, financial, manufacturing and engineering applications, are rarely static in time, and demand models able to recognize patterns in both space and time. Current machine learning (ML) and deep learning (DL) models adapted for time series processing tend to grow in …

Demonstration Of Visible And Near Infrared Raman Spectrometers And Improved Matched Filter Model For Analysis Of Combined Raman Signals, Alexander Matthew Atkinson

Electrical & Computer Engineering Theses & Dissertations

Raman spectroscopy is a powerful analysis technique that has found applications in fields such as analytical chemistry, planetary sciences, and medical diagnostics. Recent studies have shown that analysis of Raman spectral profiles can be greatly assisted by use of computational models with achievements including high accuracy pure sample classification with imbalanced data sets and detection of ideal sample deviations for pharmaceutical quality control. The adoption of automated methods is a necessary step in streamlining the analysis process as Raman hardware becomes more advanced. Due to limits in the architectures of current machine learning based Raman classification models, transfer from pure …

Laser-Spark Multicharged Ion Implantation System ‒ Application In Ion Implantation And Neural Deposition Of Carbon In Nickel (111), Oguzhan Balki

Electrical & Computer Engineering Theses & Dissertations

Carbon ions generated by ablation of a carbon target using an Nd:YAG laser pulse (wavelength λ = 1064 nm, pulse width τ = 7 ns, and laser fluence of 10-110 J/cm²) are characterized. Time-of-flight analyzer, a three-mesh retarding field analyzer, and an electrostatic ion energy analyzer are used to study the charge and energy of carbon ions generated by laser ablation. The dependencies of the ion signal on the laser fluence, laser focal point position relative to target surface, and the acceleration voltage are described. Up to C⁴⁺ are observed. When no acceleration voltage is applied between …

Pulse Power Effects On Transient Plasma Ignition For Combustion, David Wayne Alderman Ii

Electrical & Computer Engineering Theses & Dissertations

Transient plasma ignition (TPI) uses highly non-equilibrium plasmas, driven by less than 100 nanosecond, high-voltage pulses, to initiate combustion. The effects of pulse repetition frequency (PRF) and ultrashort nanosecond rise times on TPI are investigated in this work using lean, stoichiometric, and rich air-fuel mixtures at atmospheric pressure. Experimental data show the transient plasmas driven by ultrashort rise time, high voltage pulses at high PRF’s enhance the combustion of lean or stoichiometric air-methane mixtures in a static chamber. In particular, increasing PRF enhances the combustion performance by means of reduced delay times independent of the equivalence ratio of the air-fuel …

Computational Analysis Of Antipode Algorithms For The Output Feedback Hopf Algebra, Lance Berlin

Electrical & Computer Engineering Theses & Dissertations

The feedback interconnection of two systems written in terms of Chen-Fliess series can be described explicitly in terms of the antipode of the output feedback Hopf algebra. At present, there are three known computational approaches to calculating this antipode: the left coproduct method, the right coproduct method, and the derivation method. Each of these algorithms is defined recursively, and thus becomes computationally expensive quite quickly. This motivates the need for a more complete understanding of the algorithmic complexity of these methods, as well as the development of new approaches for determining the Hopf algebra antipode. The main goals of this …

Using Feature Extraction From Deep Convolutional Neural Networks For Pathological Image Analysis And Its Visual Interpretability, Wei-Wen Hsu

Electrical & Computer Engineering Theses & Dissertations

This dissertation presents a computer-aided diagnosis (CAD) system using deep learning approaches for lesion detection and classification on whole-slide images (WSIs) with breast cancer. The deep features being distinguishing in classification from the convolutional neural networks (CNN) are demonstrated in this study to provide comprehensive interpretability for the proposed CAD system using the domain knowledge in pathology. In the experiment, a total of 186 slides of WSIs were collected and classified into three categories: Non-Carcinoma, Ductal Carcinoma in Situ (DCIS), and Invasive Ductal Carcinoma (IDC). Instead of conducting pixel-wise classification (segmentation) into three classes directly, a hierarchical framework with the …

Cyber Security- A New Secured Password Generation Algorithm With Graphical Authentication And Alphanumeric Passwords Along With Encryption, Akash Rao

Electrical & Computer Engineering Theses & Dissertations

Graphical passwords are always considered as an alternative of alphanumeric passwords for their better memorability and usability [1]. Alphanumeric passwords provide an adequate amount of satisfaction, but they do not offer better memorability compared to graphical passwords [1].

On the other hand, graphical passwords are considered less secured and provide better memorability [1]. Therefore many researchers have researched on graphical passwords to overcome the vulnerability. One of the most significant weaknesses of the graphical passwords is "Shoulder Surfing Attack," which means, sneaking into a victim's computer to learn the whole password or part of password or some confidential information. Such …

On Analytic Nonlinear Input-Output Systems: Expanded Global Convergence And System Interconnections, Irina M. Winter Arboleda

Electrical & Computer Engineering Theses & Dissertations

Functional series representations of nonlinear systems first appeared in engineering in the early 1950’s. One common representation of a nonlinear input-output system are Chen-Fliess series or Fliess operators. Such operators are described by functional series indexed by words over a noncommutative alphabet. They can be viewed as a noncommutative generalization of a Taylor series. A Fliess operator is said to be globally convergent when its radius of convergence is infinite, in other words, when there is no a priori upper bound on both the L1-norm of an admissible input and the length of time over which the corresponding output is …

Generation Of Large-Volume Diffuse Plasma By An External Ionization Wave From A Single-Electrode Plasma Jet, Seyed Hamid Razavi Barzoki

Electrical & Computer Engineering Theses & Dissertations

A non-thermal transient diffuse plasma can be generated remotely in a nonconductive reduced pressure chamber by an external guided fast ionization wave (FIW). We found that an atmospheric-pressure low-temperature plasma jet (APPJ) can be a source of FIW which transfers an enhanced electric field at the wave front across a reduced pressure Pyrex glass chamber with no electrical connection to the chamber. Here, we studied the formation and propagation of the APPJ plasma, the interaction of atmospheric-pressure guided FIW with a dielectric surface which forms the wall of the reduced-pressure system, and the formation and propagation of the reduce-pressure FIW …

Non-Destructive Evaluation For Composite Material, Desalegn Temesgen Delelegn

Electrical & Computer Engineering Theses & Dissertations

The Nondestructive Evaluation Sciences Branch (NESB) at the National Aeronautics and Space Administration (NASA) Langley Research Center (LaRC) has conducted impact damage experiments over the past few years with the goal of understanding structural defects in composite materials. The Data Science Team within the NASA LaRC Office of the Chief Information Officer (OCIO) has been working with the Non-Destructive Evaluation (NDE) subject matter experts (SMEs), Dr. Cheryl Rose, from the Structural Mechanics & Concepts Branch and Dr. William Winfree, from the Research Directorate, to develop computer vision solutions using digital image processing and machine learning techniques that can help identify …

Development Of A Laser-Spark Multicharged Ion System – Application In Shallow Implantation Of Sic By Boron And Barium, Md. Haider Ali Shaim

Electrical & Computer Engineering Theses & Dissertations

A novel multicharged ion source, using laser ablation induced plasma coupled with spark discharge, has been investigated in this work. The designed and demonstrated ion source is cost-effective, compact and versatile. Experiments are described with the intention of demonstrating the practicability of ion implantation via laser ion source.

Multicharged aluminum ions are generated by a ns Q-switched Nd:YAG laser pulse ablation of an aluminum target in an ultrahigh vacuum. The experiments are conducted using laser pulse energies of 45–90 mJ focused on the Al target surface by a lens with an 80-cm focal length to 0.0024 cm² spot area …

Deep Recurrent Learning For Efficient Image Recognition Using Small Data, Mahbubul Alam

Electrical & Computer Engineering Theses & Dissertations

Recognition is fundamental yet open and challenging problem in computer vision. Recognition involves the detection and interpretation of complex shapes of objects or persons from previous encounters or knowledge. Biological systems are considered as the most powerful, robust and generalized recognition models. The recent success of learning based mathematical models known as artificial neural networks, especially deep neural networks, have propelled researchers to utilize such architectures for developing bio-inspired computational recognition models. However, the computational complexity of these models increases proportionally to the challenges posed by the recognition problem, and more importantly, these models require a large amount of data …

Computational Modeling For Abnormal Brain Tissue Segmentation, Brain Tumor Tracking, And Grading, Syed Mohammad Shamin Reza

Electrical & Computer Engineering Theses & Dissertations

This dissertation proposes novel texture feature-based computational models for quantitative analysis of abnormal tissues in two neurological disorders: brain tumor and stroke. Brain tumors are the cells with uncontrolled growth in the brain tissues and one of the major causes of death due to cancer. On the other hand, brain strokes occur due to the sudden interruption of the blood supply which damages the normal brain tissues and frequently causes death or persistent disability. Clinical management of these brain tumors and stroke lesions critically depends on robust quantitative analysis using different imaging modalities including Magnetic Resonance (MR) and Digital Pathology …

Speech Based Machine Learning Models For Emotional State Recognition And Ptsd Detection, Debrup Banerjee

Electrical & Computer Engineering Theses & Dissertations

Recognition of emotional state and diagnosis of trauma related illnesses such as posttraumatic stress disorder (PTSD) using speech signals have been active research topics over the past decade. A typical emotion recognition system consists of three components: speech segmentation, feature extraction and emotion identification. Various speech features have been developed for emotional state recognition which can be divided into three categories, namely, excitation, vocal tract and prosodic. However, the capabilities of different feature categories and advanced machine learning techniques have not been fully explored for emotion recognition and PTSD diagnosis. For PTSD assessment, clinical diagnosis through structured interviews is a …

Low Temperature Plasma For The Treatment Of Epithelial Cancer Cells, Soheila Mohades

Electrical & Computer Engineering Theses & Dissertations

Biomedical applications of low temperature plasmas (LTP) may lead to a paradigm shift in treating various diseases by conducting fundamental research on the effects of LTP on cells, tissues, organisms (plants, insects, and microorganisms). This is a rapidly growing interdisciplinary research field that involves engineering, physics, life sciences, and chemistry to find novel solutions for urgent medical needs. Effects of different LTP sources have shown the anti-tumor properties of plasma exposure; however, there are still many unknowns about the interaction of plasma with eukaryotic cells which must be elucidated in order to evaluate the practical potential of plasma in cancer …

Computational Modeling Of Facial Response For Detecting Differential Traits In Autism Spectrum Disorders, Manar D. Samad

Electrical & Computer Engineering Theses & Dissertations

This dissertation proposes novel computational modeling and computer vision methods for the analysis and discovery of differential traits in subjects with Autism Spectrum Disorders (ASD) using video and three-dimensional (3D) images of face and facial expressions. ASD is a neurodevelopmental disorder that impairs an individual’s nonverbal communication skills. This work studies ASD from the pathophysiology of facial expressions which may manifest atypical responses in the face. State-of-the-art psychophysical studies mostly employ na¨ıve human raters to visually score atypical facial responses of individuals with ASD, which may be subjective, tedious, and error prone. A few quantitative studies use intrusive sensors on …