VARIABLE-METRIC BUNDLE METHODS - FROM CONCEPTUAL TO IMPLEMENTABLE FORMS

To minimize a convex function, we combine Moreau-Yosida regularization s, quasi-Newton matrices and bundling mechanisms. First we develop con ceptual forms using ''reversal'' quasi-Newton formulae and we state th eir global and local convergence. Then, to produce implementable versi ons, we incorporate a bundle strategy together with a ''curve-search'' . No convergence results are given for the implementable versions; how ever some numerical illustrations show their good behaviour even for l arge-scale problems.