Date of Graduation

5-2025

Document Type

Thesis

Degree Name

Bachelor of Science in Data Science

Degree Level

Undergraduate

Department

Data Science

Advisor/Mentor

Sullivan, Kelly

Committee Member

Liao, Haitao

Second Committee Member

Pohl, Ed

Abstract

This honors thesis builds off work initially accepted for publication in the Proceedings of the 2025 IISE Annual Conference & Expo, which introduced “Simulation-Enhanced Bayesian Optimization” (SEBO)—a hybrid testing optimization approach that combined the usage of unbiased but costly physical experiments with the usage of cheaper but potentially biased computer experiments to optimize engineered systems. The original study established the SEBO methodology and demonstrated its effectiveness on a multimodal, two-dimensional benchmark function. Expanding on the work performed, we conduct a broader evaluation of the SEBO framework through parameter testing and experimentation under a variety of additional benchmark functions. This investigation seeks to assess the robustness, flexibility, and performance of our method under a variety of different testing conditions. Results of the study further showcase the potential of the SEBO methodology and enhance our understanding of hybrid Bayesian optimization techniques.

Keywords

Bayesian Optimization; Sequential Analysis; Physical Experiments; Computer Experiments; Black-Box Optimization

Download

Share

COinS