[Translate to English:] lundi 9 juillet, 14h, salle C 131
Mehiddin Al-Baali (Department of Mathematics and Statistics, Sultan Qaboos University, Muscat, Oman)
Recent techniques for improving the class of conjugate gradient (CG) methods for large-scale unconstrained optimization will be considered. These techniques are introduced to the methods to obtain certain useful properties of the improved CG methods. Numerical results for a selection of CG algorithms and their modifications (in particular those of Fletcher-Reeves, Polak-Ribi\’ere and Hestenes-Stiefel methods) will be described. It will be shown that the proposed techniques improve the performance of several CG algorithms substantially.