[Home]   [  News]   [  Events]   [  People]   [  Research]   [  Education]   [Visitor Info]   [UCSD Only]   [Admin]
Home > Events > CCoM > Abstract
Search this site:

Randolph E. Bank
Philip E. Gill
Michael Holst

Administrative Contact:
Terry Le

Office: AP&M 7431
Phone: (858)534-9813
Fax: (858)534-5273
E-mail: tele@ucsd.edu
Primal and Dual Active-Set Methods for Quadratic Programming

Elizabeth Wong


We present an active-set quadratic programming (QP) method based on inertia control. The method is appropriate for problems with many degrees of freedom and problems that are not necessarily convex, making it particularly useful in sequential quadratic programming (SQP) methods that use exact second derivatives. In the convex case, the method is applied to the dual QP, which may be suitable for QPs arising in mixed integer nonlinear programming, where points may be dual feasible but primal infeasible. The inertia-controlling property prevents singularity in the associated linear systems, which allows the straightforward application of modern "off-the-shelf" linear algebra software.

Tuesday, October 12, 2010
11:00AM AP&M 2402