LOCAL CONVERGENCE OF PREDICTOR-CORRECTOR INFEASIBLE-INTERIOR-POINT ALGORITHMS FOR SDPS AND SDLCPS

Authors
KOJIMA M SHIDA M SHINDOH S
Citation
M. Kojima et al., LOCAL CONVERGENCE OF PREDICTOR-CORRECTOR INFEASIBLE-INTERIOR-POINT ALGORITHMS FOR SDPS AND SDLCPS, Mathematical programming, 80(2), 1998, pp. 129-160
Citations number
35
Categorie Soggetti
Operatione Research & Management Science",Mathematics,"Computer Science Software Graphycs Programming","Operatione Research & Management Science",Mathematics,"Computer Science Software Graphycs Programming
Journal title
Mathematical programming → ACNP
ISSN journal
00255610
Volume
80
Issue
2
Year of publication
1998
Pages
129 - 160
Database
ISI
SICI code
0025-5610(1998)80:2<129:LCOPIA>2.0.ZU;2-D
Abstract
An example of an SDP (semidefinite program) exhibits a substantial dif ficulty in proving the superlinear convergence of a direct extension o f the Mizuno-Todd-Ye type predictor-corrector primal-dual interior-poi nt method for LPs (linear programs) to SDPs, and suggests that we need to force the generated sequence to converge to a solution tangentiall y to the central path (or trajectory). A Mizuno-Todd-Ye type predictor -corrector infeasible-interior-point algorithm incorporating this addi tional restriction for monotone SDLCPs (semidefinite linear complement arity problems) enjoys superlinear convergence under strict complement arity and nondegeneracy conditions. (C) 1998 The Mathematical Programm ing Society, Inc. Published by Elsevier Science B.V.