Hashing vectors for tabu search

Autor: Eitan Zemel, David L. Woodruff
Rok vydání: 1993
Předmět:
Zdroj: Annals of Operations Research. 41:123-137
ISSN: 1572-9338
0254-5330
DOI: 10.1007/bf02022565
Popis: Tabu search as proposed by Glover [3,4] has proven to be a very effective metaheuristic for hard problems. In this paper we propose that hash functions be used to record the solutions encountered during recent iterations of the search in a long list. Hash values of potential solutions can be compared to the values on the list for the purpose of avoiding cycling. This frees the algorithm designer of the need to consider cycling when creating tabu restrictions based on move attributes. We suggest specific functions that result in very good performance.
Databáze: OpenAIRE