Quantcast

Hybrid Conjugate Gradient Algorithm for Unconstrained Optimization

Research paper by N. Andrei

Indexed on: 31 Dec '08Published on: 31 Dec '08Published in: Journal of Optimization Theory and Applications



Abstract

In this paper a new hybrid conjugate gradient algorithm is proposed and analyzed. The parameter βk is computed as a convex combination of the Polak-Ribière-Polyak and the Dai-Yuan conjugate gradient algorithms, i.e. βkN=(1−θk)βkPRP+θkβkDY. The parameter θk in the convex combination is computed in such a way that the conjugacy condition is satisfied, independently of the line search. The line search uses the standard Wolfe conditions. The algorithm generates descent directions and when the iterates jam the directions satisfy the sufficient descent condition. Numerical comparisons with conjugate gradient algorithms using a set of 750 unconstrained optimization problems, some of them from the CUTE library, show that this hybrid computational scheme outperforms the known hybrid conjugate gradient algorithms.