Term of Award

Spring 2024

Degree Name

Master of Science, Information Technology

Document Type and Release Option

Thesis (open access)

Copyright Statement / License for Reuse

Digital Commons@Georgia Southern License

Department

Department of Information Technology

Committee Chair

Hayden Wimmer

Committee Member 1

Meenalosini Vimal Cruz

Committee Member 2

Jongyeop Kim

Abstract

The integration of Machine Learning (ML) and Artificial Intelligence (AI) algorithms has radically changed predictive modeling and classification tasks, enhancing a multitude of domains with unprecedented analytical capabilities. Predictive modeling leverages ML and AI to forecast future trends or behaviors based on historical data, while classification tasks categorize data into distinct classes, from email filtering to medical diagnosis. Concurrently, text-to-image generation has emerged as a transformative potential, allowing visual content creation directly from textual descriptions. These advancements are pivotal in design, art, entertainment, and visual communication, as well as enhancing creativity and productivity. This work explores three significant studies in ML and AI research, focusing on predictive and classification solutions on cloud platforms. First, a study evaluates regression-type ML models across cloud platforms, offering critical insights for optimizing models and deployment strategies. Second, research on customizing large language models for email classification addresses cybersecurity concerns, bolstering email security measures. Moreover, this work demonstrates how LLMs can be customized via training existing models on new data. Finally, investigation into text-to-image generation diffusion models highlights the evolving landscape of AI-driven visual content generation while informing future advancements and applications. Together, these studies advance the capabilities and applications of ML and AI technologies, addressing real-world challenges and driving innovation.

Research Data and Supplementary Material

No

Share

COinS