Exact penalty functions in proximal bundle methods for constrained convex nondifferentiable minimization (Q1181737)

The paper deals with nondifferentiable convex constrained optimization. Proximal bundle type methods together with exact \(\ell_ 1\) or \(\ell_ \infty\) exact penalty approaches are used. The penalty updates limit unnecessarily the penalty growth. Global convergence of the methods is established together with finite termination for polyhedral problems. Some numerical experience is reported which strongly supports the proposed ideas.