An Interior Newton Method for Quadratic Programming

Abstract

Quadratic programming represents an extremely important class of optimization problem. In this paper, we propose a new (interior) approach for the general quadratic programming problem. We establish that our new method is globally and quadratically convergent - published alternative interior approaches do not share such strong convergence properties for the nonconvex case. We also report on the results of preliminary numerical experiments: the results indicate that the proposed method has considerable practical potential.