Department of Mathematics - Seminar on Applied Mathematics - A smoothing technique for non-smooth optimization with applications to constrained minimax problems

10:00am - 11:00am
Room 4502 (near Lift 25 & 26)

The minimax problems are an important class of optimization problems arising in image processing and machine learning. Efficient algorithms for solving strongly-convex-concave type of minimax problems have been proposed. However, more general cases, in particular the constrained minimax problems are much less studied in the literature. We propose a new framework for solving constrained semi-infinite minimax problems, based on a smoothing technique for the maximum function. The convergence property and the convergence rate are presented.

Event Format
Speakers / Performers:
Prof. Ke YIN
Huazhong University of Science and Technology
Language
English
Organizer
Department of Mathematics
Post an event
Campus organizations are invited to add their events to the calendar.