We present a primal-dual interior point method for constrained nonlinear, discrete minimax problems where the objective functions and constraints are not necessarily convex. The algorithm uses two merit functions to ensure progress towards the points satisfying the first order optimality conditions of the original problem. Convergence properties are described and numerical results provided.
Dept. of Computing, Imperial College, London SW7 2AZ, UK
View An interior point algorithm for nonlinear minimax problems