Event Detail

Event Type: 
Applied Mathematics and Computation Seminar
Friday, November 20, 2020 - 12:00 to 13:00

Speaker Info

Argonne National Laboratory

Many science and engineering problems involve models, objectives, and constraints that are available only by observation or simulation, rather than by a collection of closed-form functions as is assumed in classical mathematical optimization. Consequently, derivative information with respect to decision and/or uncertain variables is often incomplete. We review algorithms for so-called derivative-free (or "zero-order") optimization and illustrate ways that mathematical structure can be exploited to significantly improve performance in a variety of real-world settings.
Joint work with Raghu Bollapragada, Wendy Di, Jeff Larson, and Matt Menickelly.

BIO: Stefan Wild is a Computational Mathematician and Director of the Laboratory for Applied Mathematics, Numerical Software, and Statistics (LANS) at Argonne National Laboratory and a Senior Fellow in the Northwestern Argonne Institute for Science and Engineering at Northwestern University. Wild obtained his Ph.D. in operations research from Cornell University and his M.S. and B.S. in applied mathematics from the University of Colorado. Wild's primary research focus is developing model-based algorithms and software for challenging numerical optimization and model calibration problems.