Engineering Commons^™

Systems And Methods For Scalable Retinal Screening Programs, Jeremy Richard Benson

Computer Science ETDs

This dissertation addresses gaps in artificial intelligence-based computer vision tasks in the medical image processing field. We demonstrate effective methods for standardizing and augmenting digital fundus photographs so that robust convolutional neural network-based systems can perform high-throughput disease classification and generalize to never-before-seen data from novel camera technologies, scaling with the changing hardware landscape, as well as keeping up with vast amount of incoming data from the ever-increasing population. We also tackle the problem of discovering relevant samples in an unlabeled cohort of image data, thus widening the bottleneck to all downstream supervised machine learning tasks.

Predicting Patient Outcomes With Machine Learning For Diverse Health Data, Dingwen Li

McKelvey School of Engineering Theses & Dissertations

As digitized clinical and health data become ubiquitous, machine learning techniques have shown promise in predicting various clinical outcomes. In this thesis research, we exploit three types of data including (1) data collected through wearables outside hospitals, (2) electronic health records (EHR) data of inpatient in general hospital wards, (3) intraoperative data collected during surgery. This thesis work investigates machine learning approaches for the diverse clinical and health data with distinctive characteristics and challenges in the context of real-world clinical applications. Specifically, this thesis makes the following contributions to the state of the art of clinical machine learning.

Extracting informative …

Quantifying And Reversing Compensatory Movements By Persons Post-Stroke In The Ambient Setting, Aaron Miller

Doctoral Dissertations

Nearly 800,000 people in the United States suffer stroke annually. Following the onset of stroke, survivors will exhibit deficits, such as hemiplegia, which will limit their function and ability to perform activities of daily living (ADLs). In order to regain independence, many stroke survivors will employ maladaptive compensatory strategies to help with the completion of tasks. Compensation is generally defined as any performance of a task that is different than the way it may have been performed before the onset of a neurodegenerative disorder. While for some severely impaired individuals, compensation may be necessary, for most these maladaptive strategies ultimately …

Deep Learning Strategies For Pool Boiling Heat Flux Prediction Using Image Sequences, Connor Heo

Graduate Theses and Dissertations

The understanding of bubble dynamics during boiling is critical to the design of advanced heater surfaces to improve the boiling heat transfer. The stochastic bubble nucleation, growth, and coalescence processes have made it challenging to obtain mechanistic models that can predict boiling heat flux based on the bubble dynamics. Traditional boiling image analysis relies on the extraction of the dominant physical quantities from the images and is thus limited to the existing knowledge of these quantities. Recently, machine-learning-aided analysis has shown success in boiling crisis detection, heat flux prediction, real-time image analysis, etc., whereas most of the existing studies are …

Modeling The Mechanical Behavior And Shock Propagation Of Metallic And Nanocomposite Materials, Pouya Shojaeishahmirzadi

UNLV Theses, Dissertations, Professional Papers, and Capstones

Evaluating the materials properties under different loading conditions is critical in various industries. Compared to quasi-static loading, predicting the behavior of structures under dynamic loads is more challenging. In this work, we will address multiple problems with strain rates varying from quasi-static to hypervelocity conditions. Computer simulation is increasingly used in the design and evaluation phases to improve the efficiency, cost-effectiveness, and flexibility. However, verification and validation of each simulation is necessary. Experiments are performed in all topics and the computational models are validated by comparing with the experiments. One of the most common types of connections in structures is …

Predicting Zero Bin In The Semiconductor Manufacturing Industry: Machine Learning Algorithms, Yazmin Montoya

Open Access Theses & Dissertations

The semiconductor industry has faced supply chain manufacturing shortages that ultimately led to a worldwide chip shortage during the COVID-19 pandemic. These chip manufacturers use sophisticated and advanced manufacturing machinery in their fabs to manufacture chips. As experienced during the pandemic, manufacturing unavailability is often due to the lack of critical manufacturing-related spare parts. This thesis evaluates the effectiveness of machine learning algorithms to identify significant factors contributing to manufacturing part outages (i.e., zero-bin) to keep manufacturing equipment running at total capacity within the organization. We propose clustering methods to segment the data and use logistic regression, logistic lasso regression, …

Advanced Analytics In Smart Manufacturing: Anomaly Detection Using Machine Learning Algorithms And Parallel Machine Scheduling Using A Genetic Algorithm, Meiling He

Theses and Dissertations

Industry 4.0 offers great opportunities to utilize advanced data processing tools by generating Big Data from a more connected and efficient data collection system. Making good use of data processing technologies, such as machine learning and optimization algorithms, will significantly contribute to better quality control, automation, and job scheduling in Smart Manufacturing. This research aims to develop a new machine learning algorithm for solving highly imbalanced data processing problems, implement both supervised and unsupervised machine learning auto-selection frameworks for detecting anomalies in smart manufacturing, and develop a genetic algorithm for optimizing job schedules on unrelated parallel machines. This research also …

A Deep Recurrent Neural Network With Iterative Optimization For Inverse Image Processing Applications, Masaki Ikuta

Theses and Dissertations

Many algorithms and methods have been proposed for inverse image processing applications, such as super-resolution, image de-noising, and image reconstruction, particularly with the recent surge of interest in machine learning and deep learning methods.

As for Computed Tomography (CT) image reconstruction, the most recently proposed methods are limited to image domain processing, where deep learning is used to learn the mapping between a true image data set and a noisy image data set in the image domain. While deep learning-based methods can produce higher quality images than conventional model-based algorithms, these methods have a limitation. Deep learning-based methods used in …

Analysis Of Deep Learning Methods For Wired Ethernet Physical Layer Security Of Operational Technology, Lucas Torlay

All Theses

The cybersecurity of power systems is jeopardized by the threat of spoofing and man-in-the-middle style attacks due to a lack of physical layer device authentication techniques for operational technology (OT) communication networks. OT networks cannot support the active probing cybersecurity methods that are popular in information technology (IT) networks. Furthermore, both active and passive scanning techniques are susceptible to medium access control (MAC) address spoofing when operating at Layer 2 of the Open Systems Interconnection (OSI) model. This thesis aims to analyze the role of deep learning in passively authenticating Ethernet devices by their communication signals. This method operates at …

Statistical Modeling, Learning And Computing For Stochastic Dynamics Of Complex Systems, Mohammadmahdi Hajiha

Graduate Theses and Dissertations

With the recent advances in sensor technology, it is much easier to collect and store streams of system operational and environmental (SOE) data. These data can be used as input to model the underlying behavior of complex engineered systems and phenomenons if appropriate algorithms with well-defined assumptions are developed. This dissertation is comprised of the research work to show the applicability of SOE data when fed into proposed tailored algorithms. The first purposes of these algorithms are to estimate and analyze the reliability of a system as elaborated in Chapter 2. This chapter provides the derivation of closed-form expressions that …

Classifying Electrocardiogram With Machine Learning Techniques, Hillal Jarrar

Master's Theses

Classifying the electrocardiogram is of clinical importance because classification can be used to diagnose patients with cardiac arrhythmias. Many industries utilize machine learning techniques that consist of feature extraction methods followed by Naive- Bayesian classification in order to detect faults within machinery. Machine learning techniques that analyze vibrational machine data in a mechanical application may be used to analyze electrical data in a physiological application. Three of the most common feature extraction methods used to prepare machine vibration data for Naive-Bayesian classification are the Fourier transform, the Hilbert transform, and the Wavelet Packet transform. Each machine learning technique consists of …

Improvement Opportunities In The Two-Source Energy Balance Model For Et Using Uav Imagery And Point Cloud Information, Mahyar Aboutalebi

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

In recent years, satellites and unmanned aerial vehicles (UAVs) provide enormous amounts of spatially-distributed information for monitoring crop conditions by measuring crop’s reflected and emitted radiation at a distance. However, applications of high-resolution UAV imagery and its intermediate products for improving crop water use estimates are not well studied. In other words, the available approaches, methods and algorithms for determining how much water to apply for irrigation using remotely sensed data have been mostly developed at satellite spatial resolutions. High-resolution imageries that have been achieved by small UAVs open new opportunities for revisiting, re-evaluating, and revising available crop water use …

Machine Learning For Unmanned Aerial System (Uas) Networking, Jian Wang

Doctoral Dissertations and Master's Theses

Fueled by the advancement of 5G new radio (5G NR), rapid development has occurred in many fields. Compared with the conventional approaches, beamforming and network slicing enable 5G NR to have ten times decrease in latency, connection density, and experienced throughput than 4G long term evolution (4G LTE). These advantages pave the way for the evolution of Cyber-physical Systems (CPS) on a large scale. The reduction of consumption, the advancement of control engineering, and the simplification of Unmanned Aircraft System (UAS) enable the UAS networking deployment on a large scale to become feasible. The UAS networking can finish multiple complex …

Numerical Modeling Of Advanced Propulsion Systems, Peetak P. Mitra

Doctoral Dissertations

Numerical modeling of advanced propulsion systems such as the Internal Combustion Engine (ICE) is of great interest to the community due to the magnitude of compute/algorithmic challenges. Fuel spray atomization, which determines the rate of fuel-air mixing, is a critical limiting process for the phenomena of combustion within ICEs. Fuel spray atomization has proven to be a formidable challenge for the state-of-the-art numerical models due to its highly transient, multi-scale, and multi-phase nature. Current models for primary atomization employ a high degree of empiricism in the form of model constants. This level of empiricism often reduces the art of predictive …

Deep Learning Approach For Dynamic Sampling For High-Throughput Nano-Desi Msi, David Helminiak

Master's Theses (2009 -)

Mass Spectrometry Imaging (MSI) extracts molecular mass data to form visualizations of molecular spatial distributions. The involved scanning procedure is conducted by moving a probe across and around a rectilinear grid, as in the case of nanoscale Desorption Electro-Spray Ionization (nano-DESI) MSI, where singular measurements can take up to ~5 seconds to acquire high-resolution (better than 10 μm) results. This temporal expense creates a high inefficiency in sample processing and throughput. For example, in a high-resolution nano-DESI study, a single mouse uterine tissue section (2.5 mm by 1.7 mm) had an acquisition time of ~4 hours to acquire 104,400 pixels. …

Rf Fingerprinting Unmanned Aerial Vehicles, Norah Ondus

Doctoral Dissertations and Master's Theses

As unmanned aerial vehicles (UAVs) continue to become more readily available, their use in civil, military, and commercial applications is growing significantly. From aerial surveillance to search-and-rescue to package delivery the use cases of UAVs are accelerating. This accelerating popularity gives rise to numerous attack possibilities for example impersonation attacks in drone-based delivery, in a UAV swarm, etc. In order to ensure drone security, in this project we propose an authentication system based on RF fingerprinting. Specifically, we extract and use the device-specific hardware impairments embedded in the transmitted RF signal to separate the identity of each UAV. To achieve …

Reaction Wheels Fault Isolation Onboard 3-Axis Controlled Satellite Using Enhanced Random Forest With Multidomain Features, Mofiyinoluwa Oluwatobi Folami

Electronic Theses and Dissertations

With the increasing number of satellite launches throughout the years, it is only natural that an interest in the safety and monitoring of these systems would increase as well. However, as a system becomes more complex it becomes difficult to generate a high-fidelity model that accurately describes all the system components. With such constraints using data-driven approaches becomes a more feasible option. One of the most commonly used actuators in spacecraft is known as the reaction wheel. If these reaction wheels are not maintained or monitored, it could result in mission failure and unwarranted costs. That is why fault detection …

Subnational Map Of Poverty Generated From Remote-Sensing Data In Africa: Using Machine Learning Models And Advanced Regression Methods For Poverty Estimation, Lionel N. Hanke

Master's Theses

According to the 2020 poverty estimates from the World Bank, it is estimated that 9.1% - 9.4% of the global population lived on less than $1.90 per day. It is estimated that the Covid-19 pandemic further aggravated the issue by pushing more than 1% of the global population below the international poverty line of $1.90 per day (WorldBank, 2020). To provide help and formulate effective measures, poverty needs to be located as exact as possible. For this purpose, it was investigated whether regression methods with aggregated remote-sensing data could be used to estimate poverty in Africa. Therefore, five distinct regression …

Data-Driven Protection Of Transformers, Phase Angle Regulators, And Transmission Lines In Interconnected Power Systems, Pallav Kumar Bera

Dissertations - ALL

This dissertation highlights the growing interest in and adoption of machine learning approaches for fault detection in modern electric power grids. Once a fault has occurred, it must be identified quickly and a variety of preventative steps must be taken to remove or insulate it. As a result, detecting, locating, and classifying faults early and accurately can improve safety and dependability while reducing downtime and hardware damage. Machine learning-based solutions and tools to carry out effective data processing and analysis to aid power system operations and decision-making are becoming preeminent with better system condition awareness and data availability.

Power transformers, …

Using Computer Vision To Track Anatomical Structures During Cochlear Implant Surgery, Nicholas Bach

McKelvey School of Engineering Theses & Dissertations

There is a steep learning curve for surgeons performing cochlear implant surgeries. We aimed to use computer vision to track anatomical features with the goal of helping surgeons perform cochlear implant surgery without damaging the cochlea. We compared nine algorithms in total, seven object tracking algorithms and two optical flow algorithms utilizing the LucasKanade method, on manually created cochlear implant surgery videos to determine the accuracy associated with each. Compared with eight other algorithms, we observed that an iterative pyramidal implementation of the Lucas-Kanade (IPLK) method, implemented through OpenCV, performed the best. The IPLK method had the lowest error rate …

Laser Surface Treatment And Laser Powder Bed Fusion Additive Manufacturing Study Using Custom Designed 3d Printer And The Application Of Machine Learning In Materials Science, Hao Wen

LSU Doctoral Dissertations

Selective Laser Melting (SLM) is a laser powder bed fusion (L-PBF) based additive manufacturing (AM) method, which uses a laser beam to melt the selected areas of the metal powder bed. A customized SLM 3D printer that can handle a small quantity of metal powders was built in the lab to achieve versatile research purposes. The hardware design, electrical diagrams, and software functions are introduced in Chapter 2. Several laser surface engineering and SLM experiments were conducted using this customized machine which showed the functionality of the machine and some prospective fields that this machine can be utilized. Chapter 3 …

Leveraging Machine Learning Techniques Towards Intelligent Networking Automation, Cesar A. Gomez

Electronic Thesis and Dissertation Repository

In this thesis, we address some of the challenges that the Intelligent Networking Automation (INA) paradigm poses. Our goal is to design schemes leveraging Machine Learning (ML) techniques to cope with situations that involve hard decision-making actions. The proposed solutions are data-driven and consist of an agent that operates at network elements such as routers, switches, or network servers. The data are gathered from realistic scenarios, either actual network deployments or emulated environments. To evaluate the enhancements that the designed schemes provide, we compare our solutions to non-intelligent ones. Additionally, we assess the trade-off between the obtained improvements and the …

Continuous-Time And Complex Growth Transforms For Analog Computing And Optimization, Oindrila Chatterjee

McKelvey School of Engineering Theses & Dissertations

Analog computing is a promising and practical candidate for solving complex computational problems involving algebraic and differential equations. At the fundamental level, an analog computing framework can be viewed as a dynamical system that evolves following fundamental physical principles, like energy minimization, to solve a computing task. Additionally, conservation laws, such as conservation of charge, energy, or mass, provide a natural way to couple and constrain spatially separated variables. Taking a cue from these observations, in this dissertation, I have explored a novel dynamical system-based computing framework that exploits naturally occurring analog conservation constraints to solve a variety of optimization …

An Anomaly Detection System For Smart Manufacturing Using Deep Learning, Tareq Tayeh

Electronic Thesis and Dissertation Repository

The smart manufacturing evolution enables financial and operational improvements across the manufacturing industry. However, smart manufacturing encompasses complex, interconnected systems which can fail at any time. To address this challenge, a novel, two-part anomaly detection system for robotic processes, with an application focus on robotic surface finishing, is presented. The first part proposes an unsupervised Attention-based Convolutional Long Short-Term Memory Autoencoder with Dynamic Thresholding (ACLAE-DT) framework for anomaly detection and diagnosis in multivariate time series of robotic surface finishing components. The second part proposes a deep residual Convolutional Neural Network-based triplet model for anomaly detection in the produced robotic surface …

Medical Image Segmentation Using Machine Learning, Masoud Khani

Theses and Dissertations

Image segmentation is the most crucial step in image processing and analysis. It can divide an image into meaningfully descriptive components or pathological structures. The result of the image division helps analyze images and classify objects. Therefore, getting the most accurate segmented image is essential, especially in medical images. Segmentation methods can be divided into three categories: manual, semiautomatic, and automatic. Manual is the most general and straightforward approach. Manual segmentation is not only time-consuming but also is imprecise. However, automatic image segmentation techniques, such as thresholding and edge detection, are not accurate in the presence of artifacts like noise …

Forecasting Pedestrian Trajectory Using Deep Learning, Arsal Syed

UNLV Theses, Dissertations, Professional Papers, and Capstones

In this dissertation we develop different methods for forecasting pedestrian trajectories. Complete understanding of pedestrian motion is essential for autonomous agents and social robots to make realistic and safe decisions. Current trajectory prediction methods rely on incorporating historic motion, scene features and social interaction to model pedestrian behaviors. Our focus is to accurately understand scene semantics to better forecast trajectories. In order to do so, we leverage semantic segmentation to encode static scene features such as walkable paths, entry/exits, static obstacles etc. We further evaluate the effectiveness of using semantic maps on different datasets and compare its performance with already …

Hardware For Quantized Mixed-Precision Deep Neural Networks, Andres Rios

Open Access Theses & Dissertations

Recently, there has been a push to perform deep learning (DL) computations on the edge rather than the cloud due to latency, network connectivity, energy consumption, and privacy issues. However, state-of-the-art deep neural networks (DNNs) require vast amounts of computational power, data, and energyâ??resources that are limited on edge devices. This limitation has brought the need to design domain-specific architectures (DSAs) that implement DL-specific hardware optimizations. Traditionally DNNs have run on 32-bit floating-point numbers; however, a body of research has shown that DNNs are surprisingly robust and do not require all 32 bits. Instead, using quantization, networks can run on …

Mitigating Insider Threat Risks In Cyber-Physical Manufacturing Systems, Jinwoo Song

Dissertations - ALL

Cyber-Physical Manufacturing System (CPMS)—a next generation manufacturing system—seamlessly integrates digital and physical domains via the internet or computer networks. It will enable drastic improvements in production flexibility, capacity, and cost-efficiency. However, enlarged connectivity and accessibility from the integration can yield unintended security concerns. The major concern arises from cyber-physical attacks, which can cause damages to the physical domain while attacks originate in the digital domain. Especially, such attacks can be performed by insiders easily but in a more critical manner: Insider Threats.

Insiders can be defined as anyone who is or has been affiliated with a system. Insiders have knowledge …

Adaptive Mobile Eeg Noise Cancellation Using 2d Convolutional Autoencoders For Bci Authentication, Tyree Lewis

USF Tampa Graduate Theses and Dissertations

Electroencephalography (EEG) signals can be used for many purposes and has the potential to be adapted to various systems. When EEG is recorded from users, these studies are performed primarily in an indoor environment, while the user is stationary. This is due to the levels of noise that are experienced when recording EEG data, to minimize errors in the data. This thesis aims to adapt tasks that are performed indoors to an external environment by removing both noise and artefacts in EEG, using a 2D Convolutional Autoencoder (CAE). The data is recorded from subjects is passed into the 2D CAE …

Studying Complex Aquifer Systems From Large-Scale Stratigraphy Development To Local Aquifer Storage And Recovery, Hamid Vahdat Aboueshagh

LSU Doctoral Dissertations

Hydrostratigraphy model is an essential component of building valid groundwater models. Many challenges are associated with constructing hydrostratigraphy models which include geological complexities such as faults, domes, and angular unconformities. Developing a method with an emphasis on capturing big data to thoroughly inform large-scale models is one of the challenges addressed in the first part of this study. The method is predicated upon discretization of the study domain into tiles based on the geological dip direction and faults. The application of the method in the state of Louisiana with the utilization of more than 114000 well logs demonstrates promising results …