Case retrieval strategies of tabu-based artificial fish swarm algorithm

Abstract

In terms of some problems existing in the process of large case base retrieval, combining tabu search method and the advantages of artificial fishschool algorithm, this paper proposes multilevel search strategy based on tabu artificial fishswarm algorithm. Tabu artificial fishswarm algorithm applies tabu table with a memory function to artificial fishswarm algorithm and uses different computing model in the similarity calculation according to properties of different types, effectively to avoid premature and blind search and other issues. Simulation results show that the algorithm outperforms other algorithms, it not only improves the retrieval accuracy and retrieval efficiency of the casebased reasoning system, but also is characterized by requiring not much with the initial values and parameters, diversity search and overcoming the local maximum, better coordinate the overall and local search capabilities and provides an effective retrieval method to retrieve the case of large case base.