一种基于定位更新技术的人工蜂群聚类算法

本文提出一种基于定位更新技术的人工蜂群算法，并将其应用于聚类分析问题. 定位更新技术是在每一次待工蜂搜索结束后，充分利用当前最优解和最差解的信息，对最优解做进一步的更新. 实验表明，基于定位更新技术的人工蜂群聚类算法，提高了算法利用先前的解来寻找更好解的开采能力. 该算法与K-means算法、基于粒子群优化的聚类算法以及基于人工蜂群的聚类算法相比，具有更好的聚类性能.
In this paper，an artificial bee colony（ABC）algorithm based on location update technology is proposed and applied to the problems of clustering analysis. The technology makes the algorithm fully use the information of current optimal solution and the worst solution to do further location update of current optimal solution after the search of onlookers. Experiments show that the ABC algorithm based on location update technology enhances the exploitation ability of applying the previous solutions to look for better solutions. The proposed algorithm also has better clustering performance compared with K-means algorithm，clustering algorithms based on particle swarm optimization and artificial bee colony