Corrector-predictor methods for monotone linear complementarity problems in a wide neighborhood of the central path

Florian A. Potra (potramath.umbc.edu)

Abstract: Two corrector-predictor interior point algorithms are proposed for solving mono\-tone linear complementarity problems. The algorithms produce a sequence of iterates in the $\caln_{\infty}^{-}$ neighborhood of the central path. The first algorithm uses line search schemes requiring the solution of higher order polynomial equations in one variable, while the line search procedures of the second algorithm can be implemented in $O(m\, n^{1+\alpha})$ arithmetic operations, where $n$ is the dimension of the problems, $\alpha\in(0,1]$ is a constant, and $m$ is the maximum order of the predictor and the corrector. If $m=\Omega(\log n)$ then both algorithms have $O(\sqrt{n}L)$ iteration complexity. They are superlinearly convergent even for degenerate problems.

Keywords: linear complementarity problem, interior-point algorithm, large neighbourhood, superlinear convergence

Category 1: Complementarity and Variational Inequalities

Category 2: Linear, Cone and Semidefinite Programming

Citation: Technical Report, UMBC, September 2004, Revised: February, 2006.

Download: [PDF]

Entry Submitted: 03/31/2006
Entry Accepted: 04/02/2006
Entry Last Modified: 03/31/2006

Modify/Update this entry