A limiting Lagrangian for infinitely constrained convex optimization inR
n |
| |
Authors: | R G Jeroslow |
| |
Institution: | (1) Georgia Institute of Technology, College of Industrial Management, Atlanta, Georgia |
| |
Abstract: | For convex optimization inR
n,we show how a minor modification of the usual Lagrangian function (unlike that of the augmented Lagrangians), plus a limiting operation, allows one to close duality gaps even in the absence of a Kuhn-Tucker vector see the introductory discussion, and see the discussion in Section 4 regarding Eq. (2)]. The cardinality of the convex constraining functions can be arbitrary (finite, countable, or uncountable).In fact, our main result (Theorem 4.3) reveals much finer detail concerning our limiting Lagrangian. There are affine minorants (for any value 0<1 of the limiting parameter ) of the given convex functions, plus an affine form nonpositive onK, for which a general linear inequality holds onR
nAfter substantial weakening, this inequality leads to the conclusions of the previous paragraph.This work is motivated by, and is a direct outgrowth of, research carried out jointly with R. J. Duffin.This research was supported by NSF Grant No. GP-37510X1 and ONR Contract No. N00014-75-C0621, NR-047-048. This paper was presented at Constructive Approaches to Mathematical Models, a symposium in honor of R. J. Duffin, Pittsburgh, Pennsylvania, 1978. The author is grateful to Professor Duffin for discussions relating to the work reported here.The author wishes to thank R. J. Duffin for reading an earlier version of this paper and making numerous suggestions for improving it, which are incorporated here. Our exposition and proofs have profited from comments of C. E. Blair and J. Borwein. |
| |
Keywords: | Lagrangians nonlinear programming Kuhn-Tucker theory convex optimization |
本文献已被 SpringerLink 等数据库收录! |
|