LANS Seminar

Seminar Title: Structured Optimal Experimental Design via Column Subset Selection for Bayesian Inverse Problems

Speaker: Hugo Diaz-Norambuena, Postdoctoral Researcher, Applied Mathematics, North Carolina State University

Date: Thursday, August 14, 2025

Time: 2:30 PM-3:30 PM (In-Person)

Location: Hybrid, Bldg. 240, Conference Room 4301

Description: We consider optimal experimental design (OED) for Bayesian inverse problems, where the experimental design variables have a certain multiway structure. Given d different experimental variables with mi choices per design variable 1 <= i <= d, the goal is to select k_i <= m_i experiments per design variable. Previous work has related OED to the column subset selection problem by mapping the design variables to the columns of a matrix A. However, this approach is applicable only to the case d=1 in which the columns can be selected independently. We develop an extension to the case where the design variables have a multi-way structure. Our approach is to map the matrix A to a tensor and perform column subset selection on mode unfoldings of the tensor. We develop an algorithmic framework with three different algorithmic templates and randomized variants of these algorithms. We analyze the computational cost of all the proposed algorithms and also develop greedy versions to facilitate comparisons. Numerical experiments on four different applications—time-dependent inverse problems, seismic tomography, X-ray tomography, and flow reconstruction—demonstrate the effectiveness and scalability of our methods for structured experimental design in Bayesian inverse problems.

Bio: Hugo Díaz-Norambuena is a postdoctoral Researcher in Applied Mathematics at North Carolina State University, working with Prof. Arvind Saibaba. I earned my M.S. and Ph.D. in Applied Mathematics from the University of Delaware under Prof. Harbir Antil and hold B.S. and M.S. degrees in Engineering from the University of Concepción, Chile, under Prof. Gabriel N. Gatica. My research spans numerical methods for PDEs, scientific computing, optimization with PDE constraints, and optimal experimental design via linear algebra, with applications in fluid dynamics, electromagnetics, and inverse problems.

Please note that the meeting URL for this event can be seen on the cels-seminars website which requires an Argonne login.