Model-Based Derivative-Free Optimization with Improved Approximation

Abstract: This talk will discuss traditional and modern approximation techniques for (expensive) derivative-free optimization, in which the approximation serves as a tool for identifying the optimality of the black-box objective, rather than merely approximating the black box itself. Static (offline) strategies and dynamic (online) strategies will be discussed. Such approximation ideas can also be naturally extended to first-order and higher-order methods.

Mathematics

Audience: researchers in the topic

PIMS-CORDS SFU Operations Research Seminar