基于混沌自适应引力搜索的模糊C均值聚类算法

摘要

针对传统模糊C均值聚类算法（FCM）易陷入局部极小值和对初值敏感的缺陷，提出一种基于混沌自适应引力搜索的模糊C均值聚类算法。首先采用自适应的更新粒子速度和混沌优化粒子最优位置的策略，对引力搜索算法进行改进。其次，用改进的引力搜索算法优化FCM的初始聚类中心。在Iris和Wine数据集上的实验表明，该算法具有很强的全局搜索能力，提高了聚类的效果和效率。%In view of the traditional fuzzy C-means(FCM)clustering algorithm the defect is easy to fall into local minimum value,and is sensitive to initial value. A kind of fuzzy C-means clustering algorithm based on chaos adaptive gravitational search was put forward. Firstly,the strategy about adaptive updating the particle velocity and the optimal position of chaos optimization particle was used to improve the gravitational search algorithm. Secondly , the improved gravitational search algorithm was used to optimize the initial clustering center of FCM. The experimental results on Iris and Wine data sets showed that the algorithm has strong global search ability ,and improves the clustering effect and efficiency.