【摘要】 为了改善聚类分析的质量,提出一种与蚁群原理相结合的聚类方法.首先对传统的聚类算法k-means进行改进,克服传统的k-means算法必须事先确定分类的个数k和选择聚类点的缺陷,然后将蚁群算法的转移概率引入k-means算法,对上述聚类结果进行二次优化.实验结果表明,改进的k-means与蚁群算法相结合的聚类方法比单一聚类算法更有效.更多还原

【Abstract】 To improve the quality of clustering analysis, the paper proposes a new clustering algorithm based on ant colony algorithm. It improves the traditional k-means algorithm, overcome the deficiency that the traditional k-means algorithm must be sure of the kinds and must select the clustering. Then the paper combines k-means algorithm with ant colony algorithm. The experimental results show that the method has a higher effect.更多还原