Variational Analysis and Optimisation Webinar, http://www.mocao.org/va-webinar/

Title: A direct search method for constrained optimization via the rounded ℓ1 penalty function.

Speaker: Christopher Price (University of Canterbury)

Date and Time: September 16th, 2020, 17:00 AEST (Register here for remote connection via Zoom)

Abstract: This talk looks at the constrained optimization problem when the objective and constraints are Lipschitz continuous black box functions. The approach uses a sequence of smoothed and offset ℓ1 penalty functions. The method generates an approximate minimizer to each penalty function, and then adjusts the offsets and other parameters. The smoothing is steadily reduced, ultimately revealing the ℓ1 exact penalty function. The method preferentially uses a discrete quasi-Newton step, backed up by a global direction search. Theoretical convergence results are given for the smooth and non-smooth cases subject to relevant conditions. Numerical results are presented on a variety of problems with non-smooth objective or constraint functions. These results show the method is effective in practice.

Author:	Vera Roshchina (via theBox)
Created:	16/09/2020
Updated:	16/09/2020
Theme:	Default
Copyright:	Creative Commons

Info