K. Beer; V. A. Skokov : Cutting plane methods based on the analytic barrier for minimization of a convex function subject to box-constraints

Author(s) :

K. Beer; V. A. Skokov

Title :

Cutting plane methods based on the analytic barrier for minimization of a convex function subject to box-constraints

Preprint series

Technische Universität Chemnitz, Fakultät für Mathematik (Germany). Preprint 99-5, 1999

Mathematics Subject Classification :

65K05 [ Mathematical programming (numerical methods) ]

90C25 [ Convex programming ]

90C06 [ Large-scale problems ]

Abstract :

Three variants of analytic barrier methods for minimization of a convex function under box constraints are analysed: the classical method of analytic barriers (centres), the proximal method of analytical barriers and the modified proximal method of analytic barriers. Theoretical and practical complexity estimates of these methods are given. Based on the numerical tests their dependence on the used control parameters is outlined. The case where the objective function is not given in an explicite analytic form and may be not defined everywhere is considered also.

Keywords :

nondifferentiable optimization, centre methods, cutting plane methods, decomposition algorithms

Language :

english

Publication time :

4/1999