Physical Sciences and Mathematics Commons^™

Utilizing Multitask Transfer Learning For Sonographic Rheumatoid Arthritis Synovitis Grading, Jordan Marie Claire Sanders

Doctoral Dissertations and Master's Theses

Classifying the four sonographic Rheumatoid Arthritis (RA) synovitis grades (Grade 0, Grade 1, Grade 2, and Grade 3) is a difficult problem due to the complexity of the relevant markers. Therefore, the current research proposes a Multitask Transfer Learning (MTL) framework for sonographic RA synovitis grading of Ultrasound (US) images in Brightness mode (B-Mode) and Power Doppler mode.

In the medical community, the lack of reliability of scoring these images has been an issue and reason for concern for doctors and other medical practitioners. The human/machine variability across the acquisition procedure of these US images creates an additional challenge that …

Online Aircraft System Identification Using A Novel Parameter Informed Reinforcement Learning Method, Nathan Schaff

Doctoral Dissertations and Master's Theses

This thesis presents the development and analysis of a novel method for training reinforcement learning neural networks for online aircraft system identification of multiple similar linear systems, such as all fixed wing aircraft. This approach, termed Parameter Informed Reinforcement Learning (PIRL), dictates that reinforcement learning neural networks should be trained using input and output trajectory/history data as is convention; however, the PIRL method also includes any known and relevant aircraft parameters, such as airspeed, altitude, center of gravity location and/or others. Through this, the PIRL Agent is better suited to identify novel/test-set aircraft.

First, the PIRL method is applied to …

Rigid Body Constrained Motion Optimization And Control On Lie Groups And Their Tangent Bundles, Brennan S. Mccann

Doctoral Dissertations and Master's Theses

Rigid body motion requires formulations where rotational and translational motion are accounted for appropriately. Two Lie groups, the special orthogonal group SO(3) and the space of quaternions H, are commonly used to represent attitude. When considering rigid body pose, that is spacecraft position and attitude, the special Euclidean group SE(3) and the space of dual quaternions DH are frequently utilized. All these groups are Lie groups and Riemannian manifolds, and these identifications have profound implications for dynamics and controls. The trajectory optimization and optimal control problem on Riemannian manifolds presents significant opportunities for theoretical development. Riemannian optimization is an attractive …

Spoken Language Processing And Modeling For Aviation Communications, Aaron Van De Brook

Doctoral Dissertations and Master's Theses

With recent advances in machine learning and deep learning technologies and the creation of larger aviation-specific corpora, applying natural language processing technologies, especially those based on transformer neural networks, to aviation communications is becoming increasingly feasible. Previous work has focused on machine learning applications to natural language processing, such as N-grams and word lattices. This thesis experiments with a process for pretraining transformer-based language models on aviation English corpora and compare the effectiveness and performance of language models transfer learned from pretrained checkpoints and those trained from their base weight initializations (trained from scratch). The results suggest that transformer language …

A System For The Detection Of Adversarial Attacks In Computer Vision Via Performance Metrics, Sarah Reynolds

Doctoral Dissertations and Master's Theses

Adversarial attacks, or attacks committed by an adversary to hijack a system, are prevalent in the deep learning tasks of computer vision and are one of the greatest threats to these models' safe and accurate use. These attacks force the trained model to misclassify an image, using pixel-level changes undetectable to the human eye. Various defenses against these attacks exist and are detailed in this work. The work of previous researchers has established that when adversarial attacks occur, different node patterns in a Deep Neural Network (DNN) are activated within the model. Additionally, it is known that CPU and GPU …

Tests Of General Relativity Through Searches For Lorentz And Cpt Symmetry Breaking, Kellie Ault

Doctoral Dissertations and Master's Theses

An effective field theory framework, the Standard Model Extension (SME), provides an agnostic, systematic test of General Relativity (GR) and its founding spacetime symmetries, Lorentz and CPT symmetry. Violating these symmetries may provide clues toward unifying the physics of the General Relativity and the Standard Model of particle physics. Part of this work involves the merge of theory, data analysis and experiments with gravitational wave (GW) signals from LIGO/Virgo/KAGRA (LVK) detectors. A modified dispersion relation derived from the SME of GWs is implemented into the LIGO Scientific Collaboration Algorithm Library Suite (LALSuite), where a joint Bayesian inference of the source …

Development And Testing Of A New Method For Velocity-Selecting White Dwarfs From Gaia By Galactic Population, Joseph Hammill

Doctoral Dissertations and Master's Theses

The detailed processes by which spiral galaxies form remains an open question in modern cosmology. Observations of the current configuration of spiral galaxies including the Milky Way reveal thin and thick disk and halo populations which must all be accounted for in formation theories and likely have distinct ages. Using the Milky Way as an example to probe this question, we are studying the formation history of these structures.

This work details our approach to age-dating the galaxy, velocity-selecting targets from a sample of white dwarfs from the Gaia DR3 catalog that have also been age-analysed using BASE-9. BASE-9 uses …

Defining Safe Training Datasets For Machine Learning Models Using Ontologies, Lynn C. Vonder Haar

Doctoral Dissertations and Master's Theses

Machine Learning (ML) models have been gaining popularity in recent years in a wide variety of domains, including safety-critical domains. While ML models have shown high accuracy in their predictions, they are still considered black boxes, meaning that developers and users do not know how the models make their decisions. While this is simply a nuisance in some domains, in safetycritical domains, this makes ML models difficult to trust. To fully utilize ML models in safetycritical domains, there needs to be a method to improve trust in their safety and accuracy without human experts checking each decision. This research proposes …

A Simulation Of The Impacts Of Climate Change On Civil Aircraft Takeoff Performance, Thomas D. Pellegrin

Doctoral Dissertations and Master's Theses

Climate change affects the near-surface environmental conditions that prevail at airports worldwide. Among these, air density and headwind speed are major determinants of takeoff performance, and their sensitivity to global warming carries potential operational and economic implications for the commercial air transport industry. Previous archival and prospective research observed a weakening in headwind strength and predicted an increase in near-surface temperatures, respectively, resulting in an increase in takeoff distances and weight restrictions. The main purpose of the present study was to update and generalize the extant prospective research using a more representative sample of worldwide airports, a wider range of …