Journal of Convex Analysis 11 (2004), No. 2, 251--266
Copyright Heldermann Verlag 2004
Identifying Active Constraints via Partial Smoothness and Prox-Regularity
W. L. Hare
Dept. of Mathematics, Simon Fraser University, Burnaby, BC V5A 1S6, Canada, firstname.lastname@example.org
A. S. Lewis
Dept. of Mathematics, Simon Fraser University, Burnaby, BC V5A 1S6, Canada, email@example.com
Active set algorithms, such as the projected gradient method in nonlinear optimization, are designed to "identify" the active constraints of the problem in a finite number of iterations. Using the notions of "partial smoothness" and "prox-regularity" we extend work of Burke, More and Wright on identifiable surfaces from the convex case to a general nonsmooth setting. We further show how this setting can be used in the study of sufficient conditions for local minimizers.
Keywords: nonlinear program, nonsmooth optimization, variational analysis, partly smooth, prox-regular, identifiable surface, projected gradient.
MSC 2000: 91C30, 49K40, 65K10.
FullText-pdf (331 KB) for subscribers only.