Department of Mathematics

Time and Place: Tuesdays 3:05-4:20 pm in Room 0231, Love Building

Course: MAP 6939

Scope: The Applied and Computational Math Seminar is series of talks with various topics covering a broad spectrum of not only applied and computational mathematics but also engineering. Researchers outside of the Department of Mathematics and Florida State University, postdocs and senior Ph.D students are also welcomed to share their work. Please contact the organizer if you wish to schedule your talk. For Spring 2023, contact Kyle Gallivan (gallivan"at"math.fsu.edu)

Abstract: Global sensitivity analysis measures the importance of the inputs of a model to its output. Several approaches, such as Sobol' sensitivity indices, derivative based global sensitivity measures (DGSM), activity scores, and Shapley values have been introduced in the literature.  In this talk, we will present a new Shapley value approach where the value function is given by the DGSM. The advantage of this approach is its computational efficiency. We will compare the new approach with two well-known Shapley value approaches, SHAP and KernelSHAP, when they are applied to problems from machine learning. The results suggest that DGSM Shapley has similar conclusions as SHAP and Kernel SHAP, but requires much less computational time.

Bio: Hui (Alyssa) Duan graduated Summa Cum Laude with honors from the Department of Mathematics from University at Buffalo, SUNY in 2017, with a Bachelor’s degree in Applied Mathematics. She obtained her Master’s degree of Computational Mathematics at Courant Institution of Mathematical Science, NYU in 2019. She is currently a 4th year Ph.D. candidate studying Financial Mathematics with advisor Dr. Okten at FSU. Her research interests include global sensitive analysis and Shapley values.