Computer Sciences Commons^™

Federated Learning Based Autoencoder Ensemble System For Malware Detection On Internet Of Things Devices, Steven Edward Arroyo

Theses and Dissertations

New technologies are being introduced at a rate faster than ever before and smaller in size. Due to the size of these devices, security is often difficult to implement. The existing solution is a firewall-segmented “IoT Network” that only limits the effect of these infected devices on other parts of the network. We propose a lightweight unsupervised hybrid-cloud ensemble anomaly detection system for malware detection. We perform transfer learning using a generalized model trained on multiple IoT device sources to learn network traffic on new devices with minimal computational resources. We further extend our proposed system to utilize federated learning …

Machine Learning Multimodal Framework For Fake News Detection And Mitigation, Nada A. Gaballah

Theses and Dissertations

Social media has become our new reality, people wake up every morning and the first thing they do before getting out of bed, is check their social media. Nowadays, people rarely read newspapers, they even rarely watch TV news or listen to radio broadcasts. In recent years, we have witnessed lots of fake news roaming social media every second, with people simply believing it and spreading it even more without checking the credibility of this news. This fake news affected several domains like what happened in the US election in 2016 and again in 2020, the false information about Covid-19 …

Case Studies For Energy Efficient Machine Learning Inference Acceleration, Recep Erol

Theses and Dissertations

The advancements in machine learning, deep learning and AI have yielded remarkable tools and innovations, but certain groups face barriers preventing their utilization of these technologies. This research identifies and categorizes these barriers, focusing on three distinct groups: those lacking computational power, seeking to deploy models across multiple devices, and struggling with optimization challenges in high-performance computing centers. The study highlights the disconnect between academia's proposed solutions and their practical integration within industries and research centers, emphasizing the lack of convenience and integration among existing tools. To bridge this gap, this research offers a multifaceted approach. Firstly, it introduces publicly …

Back To The Future: A Case For The Resurgence Of Approximation Theory For Enabling Data Driven “Intelligence”, Michael Dominic Ciocco

Theses and Dissertations

Artificial Intelligence (AI) has exploded into mainstream consciousness with commercial investments exceeding $90 billion in the last year alone. Inasmuch as consumer-facing applications such ChatGPT offer astounding access to algorithms that were hitherto restricted to academic research labs, public focus of attention on AI has created an avalanche of misinformation. The nexus of investor-driven hype, “surprising” inaccuracies in the answers provided by AI models – now anthropomorphically labeled as “hallucinations”, and impending legislation by well-meaning and concerned governments has resulted in a crisis of confidence in the science of AI. The primary driver for AI’s recent growth is the convergence …

Multi-Agent Youtube Content Discovery Bot, Ishmam Ahmed Solaiman

Theses and Dissertations

YouTube Content Discovery Bot (YTCDB) is a cutting-edge multi-agent system designed to revolutionize the video discovery process. Traditionally, researchers have faced the arduous task of manually sorting through YouTube videos to find relevant content. YTCDB leverages an analytics-driven approach to autonomously discover videos given a seed video. Each task or process within YTCDB, such as comment scraping, gathering statistics, and collecting channel data, can be efficiently handled by one or multiple agents working in tandem. This distributed approach allows for seamless coordination and delegation of tasks, ensuring optimal performance and scalability. Insights gathered from video barcoding and content analysis of …

Exporting Sysml Designs To Simulink, Drew Q. Broadbent

Theses and Dissertations

Various software systems have been developed to aid a systems engineer in evaluating system requirements, such as Dassault’s Magic System of Systems Architect (MSOSA) and MathWorks’ Simulink. Both software packages have different strengths; therefore, it is beneficial to export models from one software package to another. MSOSA provides a built-in tool that facilitates this transfer, built upon the Extension for Physical Interaction and Signal Flow Simulation (SysPhS) standard. However, the process is often unreliable and error prone and online documentation is largely lacking. This research used extensive trial and error to fill in the documentation gaps and create a method …

Customization Of Large Language Models For Causal Inference And Data Quality, Xingqiao Wang

Theses and Dissertations

In the burgeoning fields of artificial intelligence (AI) and natural language processing (NLP), Large Language Models (LLMs) have emerged as powerful tools for understanding complex textual data. This dissertation focuses on the novel customization of LLMs for enhancing causal inference in pharmacovigilance and improving entity matching for data quality—two critical challenges in healthcare analytics and data management. Through an in-depth exploration of encoder and decoder LLMs, this study illustrates how domain-specific customization can significantly advance the processing and interpretation of textual information. For pharmacovigilance, it demonstrates how tailored LLMs can extract causal relationships from adverse event reports, offering a new …

Cross-Framework Validation Of Cnn Architectures: From Pytorch To Onnx, Shreya Nandanwar

Theses and Dissertations

This research presents CIPAC (CNN Inter-framework Parameter Analysis and Comparison), a validation approach designed to ensure the integrity of Deep Learning models during their transfer between computational frameworks. Although initially tested on Convolutional Neural Networks (CNNs), CIPAC is versatile enough for various Deep Learning architectures. It goes beyond traditional methods that focus on output accuracy, by examining the models’ architecture, parameters, and components to maintain consistency after transitions, like moving from PyTorch to ONNX framework. Inspired by software architecture’s stringent validation standards, CIPAC addresses the challenges of working with Machine Learning models on different platforms, making it an essential tool …

Care-Teach: Proposing An Open-Source Approach To Personalized Learning, Jaime Augusto Alvarez Perez

Theses and Dissertations

Care-Teach is an algorithmic educational model that provides teachers and educators with the tools to create interactive, text-based lessons that address a student’s need for continuity and reinforcement. Care-Teach is built around two core components: a Student Behavior Profile and the Skill Tree. These models work together to give each student a personalized learning experience that reinforces their pre-existing strengths and inclinations. The Student Behavior Profile model keeps track of a learner’s inclinations and mood, utilizing metrics such as average response time and accuracy to categorize opportunities for educator involvement. The Skill Tree is an organizational …

Representation Learning For Generative Models With Applications To Healthcare, Astronautics, And Aviation, Van Minh Nguyen

Theses and Dissertations

This dissertation explores applications of representation learning and generative models to challenges in healthcare, astronautics, and aviation.

The first part investigates the use of Generative Adversarial Networks (GANs) to synthesize realistic electronic health record (EHR) data. An initial attempt at training a GAN on the MIMIC-IV dataset encountered stability and convergence issues, motivating a deeper study of 1-Lipschitz regularization techniques for Auxiliary Classifier GANs (AC-GANs). An extensive ablation study on the CIFAR-10 dataset found that Spectral Normalization is key for AC-GAN stability and performance, while Weight Clipping fails to converge without Spectral Normalization. Analysis of the training dynamics provided further …

Exploring Graph Neural Networks In Reinforcement Learning: A Comparative Study On Architectures For Locomotion Tasks, Gaukhar Nurbek

Theses and Dissertations

Deep Reinforcement learning (DRL) has gained importance in optimizing control policies, while Graph Neural Networks (GNNs) offer a robust approach for modeling complex relationships within systems represented as graphs. This thesis investigates the integration of DRL and GNNs to optimize control policies for robotic tasks, with a focus on locomotion. It compares static and dynamic GNN architectures for control policy predictions, revealing their strengths and limitations in adapting to locomotion predictions. The study assesses the impact of model structure complexity on GNNs' predictive capabilities, showcasing how intricate model structure can maximize GNNs' potential in capturing spatial and relational dependencies when …

Deep Learning In Indus Valley Script Digitization, Deva Munikanta Reddy Atturu

Theses and Dissertations

This research introduces ASR-net(Ancient Script Recognition), a groundbreaking system that automatically digitizes ancient Indus seals by converting them into coded text, similar to Optical Character Recognition for modern languages. ASR-net, with an 95% success rate in identifying individual symbols, aims to address the crucial need for automated techniques in deciphering the enigmatic Indus script. Initially Yolov3 is utilized to create the bounding boxes around each graphemes present in the Indus Valley Seal. In addition to that we created M-net(Mahadevan) model to encode the graphemes. Beyond digitization, the paper proposes a new research challenge called the Motif Identification Problem (MIP) related …

Space Transformation For Open Set Recognition, Atefeh Mahdavi

Theses and Dissertations

Open Set Recognition (OSR) is about dealing with unknown situations that were not learned by the models during training. In OSR, only a limited number of known classes are available at the time of training the model and the possibility of unknown classes never seen at training time emerges in the test environment. In such a setting, the unknown classes and their risk should be considered in the algorithm. Such systems require not only to identify and discriminate instances that belong to the source domain (i.e., the seen known classes contained in the training dataset) but also to reject unknown …

Denoising Diffusion Probabilistic Models Based Accelerated Mri, Alexander Francis Bugielski

Theses and Dissertations

Magnetic Resonance Imaging (MRI) is a cornerstone in obtaining intricate visualizations of anatomy and physiological processes within the human body. However, its extensive scan duration not only causes patient discomfort but also increases the likelihood of motion-induced artifacts in the images. To address such a challenge, this study investigates deep neural network models for reconstructing high-resolution MRI images from noisy and significantly undersampled data in a supervised learning manner. Specifically, it compares three models: a conventional U-Net, a self-attentive U-Net, and an innovative probabilistic diffusion model that builds upon the self-attentive U-Net architecture. These models are evaluated on their ability …

Investigating The Impact Of Human-Centered Interface Design On The User Experience Of Mobile Device Users, Ruchir Gupta

Theses and Dissertations

In order to investigate the intricate interaction between interface design, user technological proficiency, and other components of the user experience, this research study used a mixed-method approach. The beginner user group—those with little experience or expertise with technology - were the main target audience. The important discovery emphasizes the substantial influence that careful design can have on improving the effectiveness and usability of interfaces for non-tech-savvy individuals. When using the suggested Interface B instead of the current Interface A, beginner participants' task completion times significantly improved, according to the user study. This underlines the significance of creating with the needs …

Small Unmanned Aircraft System Detection And Tracking With Audio, Computer Vision, And Deep Learning Techniques, Anthony C. Brunson

Theses and Dissertations

sUAS present significant risks to local and federal agencies when under the control of negligent, reckless, or criminal operators. In the face of an escalating presence of sUAS in shared airspace with traditional aircraft, and their deployment in protected airspace as potential weapons, safeguarding personnel, facilities, and assets becomes paramount. This research seeks to address this emerging threat by investigating the efficacy of integrating low-cost distributed sensors and Machine learning (ML) models to enhance battlespace awareness and complement existing sensing platforms for real-time sUAS detection, classification, and localization. The thesis introduces the conceptualization and development of a Drone Detection Command …

Signal-To-Image Method For Counterfeit Detection In Layered Security Paradigm, Jordan Williamson

Theses and Dissertations

National level attention, resources, and priority regarding critical infrastructure have increased in recent years. This has led to adversaries and defenders exchanging positions between fortification and exploitation. One area that continues to be vulnerable is supply chain attacks like counterfeit insertion. This work investigates the application of converting collected signals into images from devices that may be considered for these critical networks. There are several aspects regarding the conversion of signals into images - specifically Red, Green, Blue (RGB) images. The methodology proposed here is potentially ideal fit for an initial security layer by achieving comparable classification results as more …

Enhancing The Resilience Of Space Systems Against Ransomware Attacks, Petersen F. Hansen

Theses and Dissertations

As their relevance has increased in recent years, space systems have become nearly essential in modern life. They are integral in the operation of navigational systems, military operations, and have ushered in a new domain of scientific inquiry. Technological advances have enabled the miniaturization of components and increased the accessibility of satellites as they find new applications in the form of Cube Satellites. However, even as these advancements have brought satellites to new heights, their interconnectedness leaves them open to new cyber threats. Ransomware attacks are one of the most prominent and disruptive cyber threats to terrestrial systems, and while …

Measuring The Presentation Of Supporting Content For A Set Of Learning Objectives Throughout The States Of An Educational Game Tree, Michael L. Hastriter Jr.

Theses and Dissertations

In an era of evolving warfare, the Department of Defense (DoD) recognizes the value of serious games as immersive tools for teaching critical concepts. This thesis introduces a pioneering framework tailored to enhance learning objectives through the presentation of educational game content. This addresses the unique needs of the DoD and other educators who use games by providing measurements to assess educational games. This researches investigates how instructors might assess games as potential teaching tools. It establishes a five-phase process, providing a framework to assess educational games against predefined learning objectives and informing future game development. This thesis demonstrates games …

Optimization Of Bluetooth Auracast Broadcast Audio Transmissions Via Signal Modeling, Lauren J. Puseman

Theses and Dissertations

Since 1994, Bluetooth has been used as a Personal Area Network (PAN) to transfer data between devices within a short range. After thirty years of progressive improvement in functionality, security, range, and power efficiency, the Bluetooth Special Interest Group has released a new feature called Auracast, which allows users to tune into nearby public audio streams and receive the feed directly to their wireless headphones or hearing aids. Soon, venues such as convention centers, museums, public forums, and sporting arenas can implement Auracast to better suit their needs. In addition to these public benefits, the Department of Defense can take …

Opportunities And Limitations: Integrating Narrative Ai Into Game-Based Assessment Creation And Evaluating Student Impacts, Kevin B. Patterson

Theses and Dissertations

The Department of Defense (DoD) has identified the need for a technically proficient workforce in the areas of science, technology, engineering, and mathematics. To meet this need, the DoD is actively seeking innovative technology capable of creating workforce development opportunities that are both accessible and effective. Educational research indicates serious games provide a potential avenue to achieve this goal. Unfortunately, a limited number of tools that simplify the game development process and leverage artificial intelligence are available. Content-generating artificial intelligence might help reduce instructor and game-based assessment designers' workloads while promoting individualized learning in students. This research presents a novel …

Malware Detection And Signature Propagation: A Study On Anti-Virus Platforms, Aaron J. Morath

Theses and Dissertations

The early detection of malware across DoD networks is paramount when considering which AV engine to employ. This study explores malware detection latency across various AV providers over a 30-day period using VirusTotal’s platform. The analysis reveals an initial surge in detections, reaching approximately 60% within 24 hours. From days 3 to 20, detections steadily increase by 1-3 instances per day, peaking at 74% on the 20th day, followed by a slight decline. The research also highlights a significant difference in false positive rates between packed and non-packed non-malicious samples, emphasizing the impact of packing on AV engine scans. While …

Enhancing Sequence With Quantum Key Distribution Protocols And An Intuitive User Interface, Blake Perkins

Theses and Dissertations

The rapidly growing domain of quantum networks necessitates advancements in associated software packages. This master’s thesis will detail, in part, new protocols added to extend the usefulness of SeQUeNCe. Notably, these added protocols were implemented to ensure compatibility and efficiency with the existing codebase. To complement this expansion in capability, the graphical user interface (GUI) was restructured. Updates to the GUI now allow users to initiate and operate these newly integrated protocols with ease, thereby expanding the accessibility of SeQUeNCe to a wider audience. By prioritizing the incorporation of these new protocols and refining the user interface, this research significantly …

Group Convolutional Decoders For Toric Codes, Jim Wang

Theses and Dissertations

Quantum Error Correction (QEC) enables both industrial and defense applications of quantum computing. Toric codes and other quantum Low-Density Parity-Check (LDPC) codes are promising and well-researched methods of QEC. However, their decoding cost increases exponentially with a computer’s qubit count. Neural Network (NN) decoders have been shown to decode a code’s error syndrome both accurately and fast enough for a real-time error correcting scheme. Recent key developments introduced Convolutional Neural Network (CNN) to implement a translationally equivariant decoder for a toric code. These CNN decoders both outperform NN decoders and require less training data. This research applies a Group Convolutional …

Modeling & Engineering Of Usmepcom Business Intelligence Data, Merrick A. Bedford

Theses and Dissertations

This thesis investigates the USMEPCOM’s issue of modeling and engineering Business Intelligence data centralized around the MEPS of Excellence (MOE) program. MEPS around the US conduct military personnel in-processing and in doing so have a vested interest in the standardization and application of the data associated with such processes. There are currently 65 MEPS stations and one RPS that handle military personnel onboarding paperwork and make determinations for military eligibility. This topic is important due to the MEPS cloud data processing system modernization efforts requiring data processing adaptation to ensure relevant and meaningful usage of current data.

Reinforcement Learning For Team Based Air Combat Maneuvering Decisions With Directed Energy Weaponry, Joshua D. Combs

Theses and Dissertations

Leveraging the Advanced Framework for Simulation, Integration, and Modeling (AFSIM) we investigate the use of reinforcement learning (RL) techniques for imbuing AUCAV agents with high-quality behaviors for the within-visual-range air combat maneuvering problem (ACMP). We formulate the 2v2 WVR ACMP as a Markov decision process wherein friendly AUCAVs are equipped with DEW capabilities and operate with 6 degrees of freedom. We utilize the Double Deep Q-Network RL algorithm, which centrally trains two friendly AUCAVs and employ a phased learning approach, initially exposing the AUCAVs to a dense reward environment for early training, followed by a sparse reward environment to encourage …

Artificial Intelligence And Perception: An Empirical Study, Anthony J. Neil

Theses and Dissertations

This thesis investigates the impact of adjusting artificial intelligence explainability levels’ outputs on user perception. The overarching study extends within the Explainable Artificial Intelligence (XAI) domain. It examines a spectrum of variables, including performance, cognizance, familiarity, transparency, system bias, and the overall impact of AI, to understand their collective and individual effects that enable effective professional use in an organization. The study aims to illuminate the relationship between the degree of explainability provided by large language models such as ChatGPT, Bard, and Bing AI and the performance of these models when tasked with XAI adjustments.

A Model-Based And System-Theoretic Approach To Design Advanced Autonomy For Air Dominance Missions: A Loyal Wingman Case Study, Elizabeth S. Pennington

Theses and Dissertations

In contested air environments, safe coordination between decision-makers is paramount. Although the Department of Defense (DoD) prioritizes the development of Artificially Intelligent (AI) wingmen for air combat, a lack of methodology exists to design safe, holistic coordination between human and autonomous wingmen in the same environment. This thesis delivers a framework using Systems Theoretic Process Analysis Extended for Coordination (STPA-Coord) to analyze and design holistic coordination for the Loyal Wingman concept in an Air Dominance mission. STPA-Coord is a safety and hazard analysis process that uses Systems Theory to analyze and design coordination between decisionmakers in a system-of-systems architecture. Using …

Sensor-Based Vehicle Classification Using Machine Learning, Luke Mcfadden

Theses and Dissertations

This research investigates the classification of vehicles into heavy and light categories using acoustic, seismic, and magnetic sensor data. The effectiveness of using frequency domain data and classical machine learning techniques, is compared with the effectiveness of using time-series data and neural networks. The primary aim in doing so was to understand if modern neural network architectures could effectively remove the need for more traditional frequency based signals processing. A significant deliverable of this thesis was the feature importance determined for each of the three phenomenological types found within the data (acoustic, seismic, and magnetic). By analyzing the importance of …

Standardization Of Risk Classifications For Unmanned Space Vehicle Missions, Collin A. Gwaltney

Theses and Dissertations

This paper seeks to model risk classification levels (A-D) for 122 Space Vehicle programs. Models include multinomial logistic regression as well as random forest, a machine learning technique based on decision trees. We use independent variables (IVs) which are theoretically correlated to risk class for the regression and one random forest model. We then include all IVs and allow the random forest technique to use those which provide the most information on risk class before paring down the number of IVs to only 7. We show that the accuracy of predictions increases from 62% to 87% by using random forest …