Optimizing Simulated Annealing Schedules with Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.4420

  author =       "Andreas Bolte and Ulrich Wilhelm Thonemann",
  title =        "Optimizing Simulated Annealing Schedules with Genetic
  journal =      "European Journal of Operational Research",
  year =         "1996",
  volume =       "92",
  number =       "2",
  pages =        "402--416",
  month =        "19 " # jul,
  keywords =     "genetic algorithms, genetic programming, Optimization,
                 Simulated annealing, Quadratic assignment problem",
  URL =          "http://www.sciencedirect.com/science/article/B6VCT-3VW8NPR-14/2/d6032805608b3a86412054ccde16f0e6",
  ISSN =         "0377-2217",
  DOI =          "doi:10.1016/0377-2217(94)00350-5",
  abstract =     "Combinatorial optimisation problems are encountered in
                 many areas of science and engineering. Most of these
                 problems are too difficult to be solved optimally, and
                 hence heuristics are used to obtain {"}good{"}
                 solutions in reasonable time. One heuristic that has
                 been successfully applied to a variety of problems is
                 Simulated Annealing. The performance of Simulated
                 Annealing depends on the appropriate choice of a key
                 parameter, the annealing schedule. Researchers usually
                 experiment with some manually created annealing
                 schedules and then use the one that performs best in
                 their algorithms.

                 This work replaces this manual search by Genetic
                 Programming, a method based on natural evolution. We
                 demonstrate the potential of this new approach by
                 optimizing the annealing schedule for a well-known
                 combinatorial optimisation problem, the Quadratic
                 Assignment Problem. We introduce two new algorithms for
                 solving the Quadratic Assignment Problem that perform
                 extremely well and outperform existing Simulated
                 Annealing algorithms.",
  notes =        "


Genetic Programming entries for Andreas Bolte Ulrich Thonemann