Research Reports from the Department of Operations

Authors

Partha Sengupta

Document Type

Dissertation

Publication Date

5-1-1981

Abstract

This research develops a nonlinear programming approach for solving a convex system of equations. Two applications of interest are also studied, namely, computation of a Brouwer fixed point of a convex function, and the Linear Complementarity Problem (LCP). Our approach to finding a zero of a function is to try to solve the optimization problem of minimizing the sum of the coordinate functions subject to each one being nonnegative. An optimal solution with objective function value of zero yields a zero of the function. The principal advantages of our approach are: (1) Global convergence of our algorithms under reasonable conditions. (2) Finite nature of our algorithm for solving piecewise linear equations and the Linear Complementarity Problem. (3) The 'descent' nature of our algorithms. By (3) we mean that with each problem we can associate a function which strictly and monotonically changes at each iteration toward its value at the solution point. This implies that at each iteration we are strictly 'improving' toward the solution. In the first part of this work we propose a finite algorithm for solving piecewise linear convex equations. We show that our algorithm finds a solution if the function satisfies a regularity condition. In the second part we transform the Linear Complementarity Problem into one of solving piecewise linear convex equations. We then show that, if the LCP has a P-matrix, then the resulting system of convex equations satisfies the needed regularity conditions. Such LCP's are therefore solved finitely by our algorithm. We also provide computational results to compare our algorithm with that of Lemke. Our algorithm seems to require greater overall computational effort than Lemke's, but it visits fewer pieces of linearity. We have identified some directions of future research which show promise for substantially improving the computational performance of our algorithm. In the third part, we use a similar approach to the problem of finding a Brouwer fixed point of a convex function mapping an n-simplex into itself. Under reasonable conditions on the function, we prove that our algorithm is globally convergent.

Keywords

Operations research, Nonlinear programming, Convex functions, Mathematical optimization, Linear complementarity problem, Mathematical analysis, Piecewise linear topology

Publication Title

Dissertation/Technical Memorandums from the Department of Operations, School of Management, Case Western Reserve University

Issue

Technical memorandum no. 493 ; Submitted in partial fulfillment of the requirements for the Degree of Doctor of Philosophy.

Rights

This work is in the public domain and may be freely downloaded for personal or academic use

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.