基于人工鱼群算法的聚类挖掘

A New Clustering Method Based on Artificial Fish-Swarm Algorithm

聚类分析就是按照数据间的相似程度,依据特定的准则将数据划分成不同种类。目前聚类分析算法普遍存在对初始参数敏感,难以找到最优聚类以及聚类有效性等问题。人工鱼群算法作为一种新型仿生优化算法,具有良好的克服局部极值和获得全局极值的能力。引入聚类数学模型,结合现有人工鱼群算法的特点和聚类算法理论,通过模拟鱼群的智能行为进行聚类分析,提出了一种基于人工鱼群算法的聚类挖掘方法。对空间数据的实验和蚁群算法的对比研究表明,该算法具有良好的聚类效果。

Clustering analysis is a division of data into similarity groups according to given rules. Traditional clustering algorithms generally have some problems, such as the sensitivity to initializing parameter, difficulty of finding out the optimized clustering result and the validity of clustering. Artificial fish - swarm algorithm（AFSA） as a novel bio - inspired optimization method possesses good capability to avoid the local extreme and obtain the global extreme. This paper introduces a mathematical model of clusters. A new clustering algorithm based on artificial fish - swarm algorithm was proposed which combined artificial fish - swarm algorithm with clustering theory based on the fundamental principle of two methods. Compared with ant colony algorithm, a well - known data mining algorithm for clustering, this method is proved to be better for clustering.