基于蚁群聚类算法的组合算法

Combinatorial Algorithm Based on Ant Colony Clustering Algorithm

基于信息熵的蚁群聚类算法是一种自组织聚类算法,具备健壮性、可视化等特点,并能生成一些新的有意义的聚类模式。基于信息素的K-means算法的K值和初始聚类中心是事先给定的,而往往两者的选择可以直接影响聚类的效果和速度(K-means算法的缺点之一)。因此,在基于信息熵的蚁群聚类算法的基础上,结合基于信息素的K-means算法,提出了一种聚类组合算法。

The information entropy based ant clustering algorithm is a self-organized clustering algorithm, which possesses the characteristics of robustness, visualization, and ability to generate some new meaningful clustering models.K values in the K- means algorithm based on the pheromone and the initial clustering center are given in advance, and often both selections can directly affect the speed of clustering effect which is the disadvantage of K-means algorithm. Therefore, a new algorithm is proposed to combine the information entropy based ant clustering algorithm with pheromone based K-means algorithm.