Monday, May 20
4:50-6:50 PM
Salon A

Semidefinite Programming (Part I of II)
(Invited Minisymposium)

Semidefinite programming (SDP) is a generalization of linear programming in which the variable space is the set of n x n real symmetric matrices, and the inequality constraint is a semidefinite matrix constraint, written X _> 0. SDP is a convex optimization problem and it has been known for some years that it enjoys much of the duality theory of LP. SDP has received much recent attention because of the realization that the interiorpoint methods developed for LP are also largely applicable to SDP, and it has been consequently widely applied in application areas ranging from combinatorial optimization to control theory. In this session, the speakers will discuss various aspects of SDP, including both theoretical and algorithmic issues.

Organizer: Michael L. Overton
Courant Institute of Mathematical Sciences, New York University

Using SDP for Non-Lipschitz Eigenvalue Optimization
James V. Burke, University of Washington; and Michael L. Overton, Organizer
Exploiting Sparsity in Interior-Point Methods for Semidefinite Programming
Lieven Vandenberghe, Stanford University
The Analytical Center, Optimal Face and Convergence of the Central Path of a
Semidefinite Program
Katya Scheinberg, Columbia University
Exact Duality Theories for Semidefinite Programming
Motakuri Ramana, University of Florida

Registration | Hotel Information | Transportation | Speaker Index | Program Overview

LMH, 3/15/96