Regularized optimization methods for convex MINLP problems

• Welington de Oliveira ^[1]
  1. [1] Universidade do Estado do Rio de Janeiro

     Universidade do Estado do Rio de Janeiro

     Brasil

     
• Localización: Top, ISSN-e 1863-8279, ISSN 1134-5764, Vol. 24, Nº. 3, 2016, págs. 665-692
• Idioma: inglés
• DOI: 10.1007/s11750-016-0413-4
• Enlaces
  • Texto completo (pdf)
• Resumen

  • We propose regularized cutting-plane methods for solving mixed-integer nonlinear programming problems with nonsmooth convex objective and constraint functions. The given methods iteratively search for trial points in certain localizer sets, constructed by employing linearizations of the involved functions. New trial points can be chosen in several ways; for instance, by minimizing a regularized cutting-plane model if functions are costly. When dealing with hard-to-evaluate functions, the goal is to solve the optimization problem by performing as few function evaluations as possible. Numerical experiments comparing the proposed algorithms with classical methods in this area show the effectiveness of our approach.