Engineering Commons^™

Asset Cueing Nuclear Radiation Anomaly Detection Using An Embedded Neural Network Resource, April Inamura

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

Nuclear radiation detection is inherently a challenging task, coupled with a high background variation or increase in anomalies, the accuracy for detection can plummet. A key factor in the success of nuclear detection hinges on the sensor’s ability to generalize its model and directly leads to the model’s robustness. The goal of this project is to develop algorithms suitable for use on the University of Nebraska-Lincoln’s Pingora chip, a low-power, system-on-chip device with an active neural processing unit (NPU) made for nuclear radiation detection. The thesis aims to improve Pingora’s overall generalization ability in nuclear radiation source detection. A multiphase …

A Data-Driven Method For Damage Detection In An Open Deck Steel Truss Railroad Bridge Under The Moving Train Load, Mahesh Pokhrel

Theses and Dissertations

Open-Deck (OD) steel truss railroad bridges are one of the most common types of railroad bridges in the United States. They are, however, significantly vulnerable to the dynamic effects of moving train load. In the long run, these dynamic effects cause fatigue damage to the structural members. The damage in most of the railroad bridges are monitored by visual inspections which are sometimes unreliable and inconsistent due to human error. This study explores the identification and classification of damage to an open deck railroad bridge from both time-domain (statistical features) and time-frequency domain features (Hilbert-Huang Transform) extracted through acceleration response …

Leveraging Programmable Switches To Enhance The Performance Of Networks: Active And Passive Deployments, Elie Kfoury

Theses and Dissertations

The performance of networks today is drastically affected by: 1) switches equipped with large buffers, referred to as “bloated buffers”: due to the lack of programmability and traffic visibility in legacy switches, operators nowadays configure large buffers statically without considering the characteristics or dynamics of flows. Such buffers increase the delays on packets, causing the Quality of Service (QoS) of networked applications (e.g., voice over IP, web browsing) to degrade; 2) switches forwarding packets on a best-effort basis: traffic crossing a switch is heterogeneous in many ways. Mixing such traffic in a single queue without any QoS measures can drastically …

Extending The Convolution In Graph Neural Networks To Solve Materials Science And Node Classification Problems, Steph-Yves Mike Louis

Theses and Dissertations

The usage of graph to represent one's data in machine learning has grown in popularity in both academia and the industry due to its inherent benefits. With its flexible nature and immediate translation to real life observed objects, graph representation had a considerable contribution in advancing the state-of-the-art performance of machine learning in materials.

In this dissertation proposal, we discuss how machines can learn from graph encoded data and provide excellent results through graph neural networks (GNN). Notably, we focus our adaptation of graph neural networks on three tasks: predicting crystal materials properties, nullifying the negative impact of inferior graph …

Predicting Material Structures And Properties Using Deep Learning And Machine Learning Algorithms, Yuqi Song

Theses and Dissertations

Discovering new materials and understanding their crystal structures and chemical properties are critical tasks in the material sciences. Although computational methodologies such as Density Functional Theory (DFT), provide a convenient means for calculating certain properties of materials or predicting crystal structures when combined with search algorithms, DFT is computationally too demanding for structure prediction and property calculation for most material families, especially for those materials with a large number of atoms. This dissertation aims to address this limitation by developing novel deep learning and machine learning algorithms for effective prediction of material crystal structures and properties. Our data-driven machine learning …

Exploratory Data-Driven Models For Water Quality: A Case Study For Tampa Bay Water, Sandra Sekyere

USF Tampa Graduate Theses and Dissertations

Water, a crucial resource for sustaining life, covers approximately 70% of the earth's surface. Nonetheless, the quality of water is deteriorating rapidly due to the rapid growth of urban areas and industries, which is a worrying trend causing harm to human health and the ecosystem. Water quality forecasting has a key role in water resources management by enabling effective pollution control, ecosystem monitoring, and decision-making.

Previously, traditional statistical models were used to forecast water quality, but they were unable to examine the non-linear relationships between water quality parameters, and they assumed that all datasets were distributed normally. This study uses …

Insect Classification And Explainability From Image Data Via Deep Learning Techniques, Tanvir Hossain Bhuiyan

USF Tampa Graduate Theses and Dissertations

Since the dawn of the Industrial Revolution, humanity has always tried to make labor more efficient and automated, and this trend is only continuing in the modern digital age. With the advent of artificial intelligence (AI) techniques in the latter part of the 20th century, the speed and scale with which AI has been leveraged to automate tasks defy human imagination. Many people deeply entrenched in the technology field are genuinely intrigued and concerned about how AI may change many of the ways in which humans have been living for millennia. Only time will provide the answers. This dissertation is …

Neural Compression Inference Accelerator: A Cost And Energy-Effective Alternative To Conventional Machine Learning Inference Methods, Matthew Wallace

Master's Theses

Recent developments in machine learning and artificial intelligence have sparked an influx of workloads that require specialized computer hardware for cloud services. The hardware running machine learning models predominantly consists of graphics processing units (GPUs) and tensor processing units (TPUs). However, these com- ponents are expensive for cloud services to purchase, costly for customers to rent, prone to price spikes, and energy-intensive. In this research we show that both cloud services and customers would benefit from utilizing field programmable gate arrays (FPGAs) to alleviate the aforementioned challenges. An FPGA can be configured as a machine learning accelerator, operating similarly to …

Enhancing Telecom Churn Prediction: Adaboost With Oversampling And Recursive Feature Elimination Approach, Long Dinh Tran

Master's Theses

Churn prediction is a critical task for businesses to retain their valuable customers. This paper presents a comprehensive study of churn prediction in the telecom sector using 15 approaches, including popular algorithms such as Logistic Regression, Support Vector Machine, Decision Tree, Random Forest, and AdaBoost.

The study is segmented into three sets of experiments, each focusing on a different approach to building the churn prediction model. The model is constructed using the original training set in the first set of experiments. The second set involves oversampling the training set to address the issue of imbalanced data. Lastly, the third set …

Neural Correlates Of Post-Traumatic Brain Injury (Tbi) Attention Deficits In Children, Meng Cao

Dissertations

Traumatic brain injury (TBI) in children is a major public health concern worldwide. Attention deficits are among the most common neurocognitive and behavioral consequences in children post-TBI which have significant negative impacts on their educational and social outcomes and compromise the quality of their lives. However, there is a paucity of evidence to guide the optimal treatment strategies of attention deficit related symptoms in children post-TBI due to the lack of understanding regarding its neurobiological substrate. Thus, it is critical to understand the neural mechanisms associated with TBI-induced attention deficits in children so that more refined and tailored strategies can …

Towards A Mobile Ad-Hoc Mesh Network Establishing Emergency Drone System, Ryan Integlia

Florida Conference on Recent Advances in Robotics

A drone based mobile ad-hoc mesh network for emergency communications is discussed. The in-progress project seeks to improve emergency and disaster area communication systems by creating a mobile, ad-hoc wireless network with an array of microcomputers, a GPS receiver, IMU, network adapter and drone. The Linux based platform includes network management, data collection, and integration with visualization. The expected outcome of this project is the establishment of a wireless mesh network capable of self-healing to support emergency response.

Exploring Machine Learning In Deep Foundation And Soil Classification Application, Mohammad Moontakim Shoaib

LSU Master's Theses

The applicability of several Machine Learning (ML) models was explored in this research to predict the ultimate capacity and load-settlement behavior of axially loaded single-driven piles from Cone Penetration Test (CPT) data. Additionally, a common CPT-based soil behavior type (SBT) classification system was reproduced using those ML models. Eighty static pile load tests and corresponding CPT data close to those pile locations were collected from 34 sites in Louisiana for the deep foundation application. On the other hand, 70 CPT soundings were taken in 14 different parishes across Louisiana for the soil classification application. Specifically, tree-based ML models such as …

Eddy Current Defect Response Analysis Using Sum Of Gaussian Methods, James William Earnest

Theses and Dissertations

This dissertation is a study of methods to automatedly detect and produce approximations of eddy current differential coil defect signatures in terms of a summed collection of Gaussian functions (SoG). Datasets consisting of varying material, defect size, inspection frequency, and coil diameter were investigated. Dimensionally reduced representations of the defect responses were obtained utilizing common existing reduction methods and novel enhancements to them utilizing SoG Representations. Efficacy of the SoG enhanced representations were studied utilizing common Machine Learning (ML) interpretable classifier designs with the SoG representations indicating significant improvement of common analysis metrics.

Utilization Of Machine Learning To Investigate Material State, Ana B. Abadie

Honors College Theses

The ability to predict material behavior that undergoes various loading conditions is critical to the development of reliable and safe components. Thermal and mechanical fatigue loading can cause significant damage to materials, leading to failure and potential safety hazards. Machine learning algorithms have emerged as a promising tool for improving accuracy and efficiency of predicting material behavior under such loading conditions. This research provides a comprehensive overview of a machine learning algorithm that is able to analyze and predict material state independently of the loading sequence. Unidirectional carbon fiber reinforced polymer (UD CFRP) composite which has undergone two different loading …

Detecting Pathobiomes Using Machine Learning, Valerie Jackson, Valerie Jackson

Industrial Engineering Undergraduate Honors Theses

Machine learning is a field with high growth potential due to the overall continuous progressions, developments, advancements, and improvements caused by the way it is used to help interpret and use large amounts of data [1]. One type of data that can be collected and analyzed by these machine learning models is data that is associated with DNA and information that the DNA gives. The research will be focusing specifically on using machine learning technology to detect pathobiomes indicative of salmonella pork. The pathobiome associated with salmonella is very similar to others, and this causes a problem for classification/detection with …

Lignin Copolymer Property Prediction Using Machine Learning, Collin Larsen

Chemical Engineering Undergraduate Honors Theses

Lignin, an abundant biopolymer, is a waste byproduct of the paper and pulp industry. Despite its renewable nature and potential applicability in various products, such as plastics and composites, the development of lignin-based materials has been impeded by the cumbersome, Edisonian process of trial and error. This research proposes a novel approach to forecasting the properties of lignin-based copolymers by utilizing a recurrent neural network (RNN) based on the Keras models previously created by Tao et al. Example units of modified lignin were synthesized via esterification and amination functional group modifications. To increase the efficiency and accuracy of the prediction …

Probabilistic Machine Learning For Battery State Of Health Prognostics, Charli Zaretsky

Honors Scholar Theses

The ability to understand and predict the state of health (SOH) of lithium-ion batteries is an integral component of their widespread commercial use. There are various methods through which SOH can be analyzed and predicted, and this paper discusses these different methods, and the strengths and weaknesses of each. This paper also details an analysis of lithium-ion battery SOH through two data-driven machine learning methods: XGBoost and Gaussian process regression. A comparison is made between each method’s accuracy in predicting next-cycle discharge capacity using electrochemical impedance spectroscopy (EIS) readings and battery charge and discharge rates, from a dataset given in …

Improving Classification In Single And Multi-View Images, Hadi Kanaan Hadi Salman

Graduate Theses and Dissertations

Image classification is a sub-field of computer vision that focuses on identifying objects within digital images. In order to improve image classification we must address the following areas of improvement: 1) Single and Multi-View data quality using data pre-processing techniques. 2) Enhancing deep feature learning to extract alternative representation of the data. 3) Improving decision or prediction of labels. This dissertation presents a series of four published papers that explore different improvements of image classification. In our first paper, we explore the Siamese network architecture to create a Convolution Neural Network based similarity metric. We learn the priority features that …

Computational Analysis Of Water Braking Phenomena For High-Speed Sled And Its Machine Learning Framework, Jose A. Terrazas

Open Access Theses & Dissertations

Specializing in high-speed testing, Holloman High-Speed Test Track (HHSTT) uses a process called "water braking" as a method to bring vehicles at the test track to a stop. This method takes advantage of the higher density of water, compared to air, to increase braking capability through momentum exchange. By studying water braking using Computational Fluid Dynamics (CFD), forces acting on track vehicles can be approximated and prepared for prior to the actual test. In this study, focus will be made on the brake component of the track sled that is responsible for interacting with the water for braking. By discretizing …

Development Of A Cost-Constrained Intelligent Prosthetic Knee With Real-Time Machine Learning, Predictive Stumble Control, Lucas Jonathan Galey

Open Access Theses & Dissertations

The field of biomechatronics is evolving quickly with advances in computer science, biology, and electrical and mechanical engineering. Coupled with increased interests in machine learning (ML) across all industry sectors, there are opportunities to leverage advanced analytics in uniquely complex problems. This study aimed to deploy real-time ML predictions in a novel microprocessor-controlled prosthetic knee (MPK) device capable of identifying and responding to stumble-events to reduce amputee fall prevalence. Innately, stumbling is a chaotic event. Current MPKs operate by detecting gait characteristics and reacting to preprogrammed states. While these systems are beneficial in significant ways, such as energy expenditure and …

Achieving Causal Fairness In Recommendation, Wen Huang

Graduate Theses and Dissertations

Recommender systems provide personalized services for users seeking information and play an increasingly important role in online applications. While most research papers focus on inventing machine learning algorithms to fit user behavior data and maximizing predictive performance in recommendation, it is also very important to develop fairness-aware machine learning algorithms such that the decisions made by them are not only accurate but also meet desired fairness requirements. In personalized recommendation, although there are many works focusing on fairness and discrimination, how to achieve user-side fairness in bandit recommendation from a causal perspective still remains a challenging task. Besides, the deployed …

Accounting And Financial Statements Auto Analysis System, Zhen Jia

Electronic Theses, Projects, and Dissertations

This project was motivated by the need to revolutionize the generation of financial statements and financial analysis process thus speeding up business decision making. The research questions were: 1) How can machine learning increase the speed of financial statement preparation and automate financial statements analysis? 2) How can businesses balance the benefits of automating financial analysis with potential concerns around privacy, data security, and bias? 3) Can the Java J2EE framework provide a reliable running environment for machine learning?

The findings were: 1) Machine learning can significantly increase the accuracy and speed of financial analysis. Using machine learning algorithms, financial …

Meat Quality Prediction Using Machine Learning, Rohit Buddiga

Electronic Theses, Projects, and Dissertations

Meat quality is an essential aspect of the food industry. However, traditional methods of meat quality prediction have limitations in terms of accuracy, cost, and time efficiency. This project focused on utilizing advanced Deep learning and Machine learning algorithms to develop- machine learning models that could predict the freshness (or spoilage) of meat with a 100% accuracy, based on image data. In addition to accuracy, this study emphasizes the significance of speed and time in selecting the optimal machine learning model. The research questions are: Q1. What hybrid neural networks should be used to predict freshness? Q2. How do hybrid …

From Point Estimates To Predictive Distributions In Machine Learning Models - A Statistical Importance Sampling Framework, Giuseppina Carannante

Theses and Dissertations

In this thesis, we leverage powerful statistical frameworks for optimal sequential estimation and tracking in non-linear and non-Gaussian dynamical models, which enjoy proven (asymptotic) optimality properties. Initially, we build upon our previous work, which employed first-order Taylor series approximation to propagate the first two predictive moments, to derive Bayesian encoder-decoder networks. This work introduced the notion of dense, pixel-level uncertainty map that is crucial in fields, such as autonomous vehicles and medical segmentation. We then extended the Bayesian framework to an ensembling scheme based on ensemble Kalman Filtering (EnKF). While EnKF represents the predictive distribution with an ensemble of draws, …

Air Conditioner Fail Safe Detector, Jonathan Li

Electrical and Computer Engineering Senior Theses

Air Conditioners are essential to human life. In an age of sudden temperature changes, moving systems, particularly HVAC (Heat Ventilation and Air Conditioning) systems, are the primary source to physically and financially protect the health of all workers, employees, and students. Air Conditioners are prone to mechanical and electrical malfunction/breakdown due to excessive use. Regular maintenance and service intervals are helpful but do not guarantee free-malfunction systems. When the systems break down, especially commercial systems, the repair cost can be highly expensive and time-consuming. Can we detect early signs of potential problems in the systems to minimize the repair cost …

A Hybrid Active Learning Framework For Personal Thermal Comfort Models, Duygu Zeynep Tekler, Yue Lei, Yuzhen Peng, Clayton Miller, Adrian Chong

Research Collection College of Integrative Studies

Personal thermal comfort models are used to predict individual-level thermal comfort responses to inform design and control decisions of buildings to achieve optimal conditioning for improved comfort and energy efficiency. However, the development of data-driven thermal comfort models requires collecting a large amount of sensor-related measurements and user-labelled data (i.e., user feedback) to achieve accurate predictions, which can be highly intrusive and labour intensive in real-world applications. In this work, we propose a hybrid active learning framework to reduce data collection costs for developing data-efficient and robust personal comfort models that predict users’ thermal comfort and air movement preferences. Through …

Estimating Crop Stomatal Conductance Through High-Throughput Plant Phenotyping, Junxiao Zhang

Department of Agricultural and Biological Systems Engineering: Dissertations, Theses, and Student Research

During photosynthesis and transpiration, crops exchange carbon dioxide and water with the atmosphere through stomata. When a crop experiences water stress, stomata are closed to reducing water loss. However, the closing of stomata also negatively affects the photosynthetic efficiency of the crop and leads to lower yields. Stomatal conductance (gs) quantifies the degree of stomatal opening and closing by using the rate of gas exchange between the crop and the atmosphere, which helps to understand the water status of the crop for better irrigation management. Unfortunately, gs measurement typically requires contact measuring instruments and manual collection in the field, which …

Getting A Handle On Floor Plan Analysis - Door Classification In Floor Plans And A Survey On Existing Datasets, João David, António Leitão

Architecture and Planning Journal (APJ)

Floor plan interpretation and reconstruction is crucial to enable the transformation of drawings to 3D models or different digital formats. It has recently taken advantage of neural-based architectures, especially in the semantic segmentation field. These techniques perform better than traditional methods, but the results depend mainly on the data used to train the networks, which is often crafted for the specific task being performed, making it hard to reuse for different purposes. In this paper, we conduct a literature survey on the existing datasets for floor plan analysis, and we explore how information regarding door placement and orientation can be …

Gestural Design - Hand Tracking For Digital Drawing, Jordan A. Kanter, Kamil Quinteros

Architecture and Planning Journal (APJ)

Computational design is increasingly interested in the active feedback between the user/designer and the digital space. Often, our initial instinct as designers comes from a gesture, a movement of the hands that gets translated into sketches and 3D models via the tools available to us. While the physical realm allows for muscle memory, tactile feedback, and creative output via movement, digital design often negates the body of the designer as it sequesters us into a screen-mouse-hand relationship. Moreover, current CAD software tools often reinforce this standardization, further limiting the potential of physical bodily gestures as a vehicle for architectural form-making. …

Generating Mass Housing Plans Through Gans - A Case In Toki, Turkey, Gi̇zem Özerol Özman, Semra Arslan Selçuk

Architecture and Planning Journal (APJ)

Nowadays, Machine Learning (ML) is frequently used in almost all disciplines having an intersection with technology. Recently, architects are using existing plan data sets in architecture through Deep Learning (DL) algorithms of big data to achieve generative and non-existent plan models by using ML. Especially, Generative Adversarial Neural Networks (GANs), one of the deep learning algorithms, have been in use in the creation of generative models for architectural studies. Within the scope of this paper, architectural drawings were generated by using GANs. This generation method allows for the training of spatial layout planning to networks and for the generation of …