基于上下文人工蜂群的模糊C均值聚类算法

Fuzzy C means clustering algorithm based on context artificial bee colony

针对传统模糊C均值聚类算法(FCM)过度依赖初始中心且易陷入局部最优等问题,提出一种基于上下文人工蜂群的模糊C均值聚类算法(CABCFCM)。首先,引入人工蜂群算法,用来确定FCM算法的初始聚类中心;其次,采用邻域半径和高斯扰动提升人工蜂群算法的局部搜索能力,并引入上下文多臂赌博机提升算法开发能力;最后,将CABCFCM应用到广告分发业务的推荐模型上。仿真实验结果表明,其准确率明显更高、聚类效果更佳。

In view of the problem that the traditional fuzzy C means clustering algorithm(FCM)relies too much on the initial center and easily falls into local optimization,a fuzzy C means clustering algorithm based on contextual artificial bee colony(CABCFCM)is proposed.Firstly,artificial bee colony algorithm is introduced to determine the initial clustering center of FCM.Secondly,the neighborhood radius and Gaussian perturbation are employed to improve the local search ability of the artificial bee colony algorithm,and a linear upper confidence bound is used to improve the algorithm exploration ability.Finally,the CABCFCM algorithm is applied to the recommendation model for advertisement business distribution,and the simulation results demonstrate that the recommendation accuracy is significantly improved and the clustering effect is much better.