University of Washington
Convex composite optimization is a broad framework for modeling a with variety of optimization problems. In this talk we introduce this framework, illustrate how it is applied in modeling, and discuss a number of its theoretical properties from exact penalization, to a multiplier theory, to augmented Lagrangians and beyond. We begin with the convex case, and then show how it can be extended to the nonconvex case. If time permits, I will also discuss algorithms.
University of Washington