A Subgradient Algorithm for Certain Minimax and Minisum Problems—The Constrained Case

摘要: We present an implementable feasible direction subgradient algorithm for minimizing the maximum of a finite collection functions subject to constraints. It is assumed that each function involved in defining objective sum basic convex and number different sets associated with nondifferentiable points on any bounded set. Problems involving $l_p$-norms, such as location approximation problems, can be put this form. Conditions are given which guarantee generates sequence converging optimal solution. The results computational tests some problems included. In these we explore sensitivity its parameters.