A Clustering Algorithm Based on Multi-agent Meta-heuristic Architecture

Zhu Qiang, Shun Yuqiang, Cen Yang
2014 International Journal of Hybrid Information Technology  
A clustering algorithm is proposed in this paper, which is based on discussion of multiagent meta-heuristic architecture of the ant colony optimization algorithm. The multi-agent architecture of ant colony optimization meta-heuristic includes three levels. Level-0 agents build solutions, level-l agents improve solutions and level-2 agents update pheromone matrix. The updated pheromone then provides feedback information for the next iteration of solution construction. Mutation probability p and
more » ... heromone resistance ρ are the adaptive parameters, which can be adjusted automatically during the evolution progress. With the adaptive variable, the algorithm can solve the contradiction between convergence speed and precocity and stagnation. The algorithm has been tested and compared with the clustering algorithm based on Genetic and Simulate annealing. Experimental results show that the proposed algorithm is more effective, and the clustering quality and efficiency are promising.
doi:10.14257/ijhit.2014.7.2.21 fatcat:5x3pjv2livcxzfefk3ejmv77be