基于模拟退火的样本加权FCM算法  

Sample weighted fuzzy C-means algorithm based on simulated annealing

作　　者：段林珊[1,2] 刘培玉[1,2] 谢方方[1,2] 

机构地区：[1]山东师范大学信息科学与工程学院,山东济南250014 [2]山东省分布式计算机软件新技术重点实验室,山东济南250014

出　　处：《计算机工程与设计》2013年第6期2004-2008,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(60873247);山东省自然科学基金项目(ZR2009GZ007;ZR2011FM030)

摘　　要：为了解决模糊C均值聚类算法(FCM)中聚类类数初始值是由先验知识人为确定并且目标函数忽略了样本属性数据之间的不均衡性问题,提出了一种基于模拟退火的样本加权FCM算法(SASWFCM),利用模拟退火算法可以寻求全局最优解的特点,计算出聚类数初始值,并对聚类中心和目标函数进行加权处理。通过实验分析,该算法与原FCM算法相比较而言,无需人为确定聚类初始值并且在分类准确数和准确率上有所提高,体现了算法的优越性,验证了改进后算法的实际价值。When using fuzzy C-means clustering algorithm, the initial value of the number of clustering classes needed to be determined by prior knowledge of people and the objective function ignored the disequilibrium problems in sample attribute data. This paper put forward a kind of sample weighted FCM algorithm based on simulated annealing algorithm. It calculated the number of clustering initial value using the simulated annealing algorithm which has an excellent feature of solving globally optimal solution problems and pressed weighted processing on the clustering center and the objective function. The experimental analysis shows that this algorithm needs not determine the number of clustering initial value by prior knowledge of people compared to other FCM algorithms and it is superior to the traditional FCM algorithm at classification accuracy and classification accuracy rate.

关 键 词：模糊C均值聚类 模拟退火 样本加权 初始聚类数 加权目标函数 自动推理 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]