多路径高斯核模糊C均值聚类算法  

A multi-path Gaussian kernel fuzzy C means clustering algorithm

作　　者：文传军[1] 汪庆淼[2] WEN Chuan-jun;WANG Qing-miao(School Mathematical Sciences and Chemical Engineering,Changzhou Institute of Technology,Changzhou 213002;School of Computer Science and Technology,Soochow University,Suzhou 215021,China)

机构地区：[1]常州工学院数理与化工学院,江苏常州213002 [2]苏州大学计算机学院,江苏苏州215021

出　　处：《计算机工程与科学》2018年第5期931-937,共7页Computer Engineering & Science

基　　金：国家自然科学基金(61170126);常州工学院校级课题(YN1305)

摘　　要：聚类算法单一迭代路径限制了参数优值的搜索。提出一种多路径高斯核模糊C均值聚类算法(MGKFCMs),MGKFCMs算法首先取核目标函数及模糊隶属度函数中的核函数为高斯核函数;然后利用梯度法得到聚类中心迭代公式,并基于该迭代公式和粒子群算法作聚类中心的并行参数迭代,在每一次聚类迭代时,选择聚类目标函数值小的路径作为参数迭代最终路径。对比分析了MGKFCMs算法的相关性质,通过仿真实验验证了所提算法的有效性。The single iteration path of the clustering algorithm limits the search path of the parame ter＇s optimal value. In this paper, a multi path Gaussian kernel fuzzy c means clustering method is pro posed and named MGKFCMs. Firstly, MGKFCMs takes the nuclear objective function and the fuzzy membership degree function in the kernel function as the Gaussian kernel function. Secondly, the gradient method is used to get the iterative formula of the clustering center. Based on this iterative formula particle swarm tively in parallel. I selected as the fina the convergence fective. algorithm, the parameters of the clustering center are calculated iteran every iteration of clustering, a path with 1 path of parameter iteration. The correlatio the algorithm is studied. Simulation result small clus n property of MGKFCMs s show that the proposed function value is is analyzed, and algorithm is effective.

关 键 词：核方法 模糊聚类 高斯核 聚类中心 多路径迭代 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]