| 一种新的模糊C均值聚类算法的优化方法 |
| |
| 引用本文： | 储慧琳,赵博,张兴明.一种新的模糊C均值聚类算法的优化方法[J].信息工程大学学报,2011,12(3):347-351. |
| |
| 作者姓名： | 储慧琳 赵博 张兴明 |
| |
| 作者单位： | 信息工程大学国家数字交换系统工程技术研究中心,河南郑州,450002 |
| |
| 基金项目： | 国家863计划资助项目 |
| |
| 摘 要： | 针对模糊C均值聚类算法对初始值敏感、易陷入局部最优的缺陷,提出一种新的优化方法.该方法通过分集聚类的结果初始化粒子群,得到逼近全局最优的聚类中心,再进行全局聚类,能有效避免陷入局部最优.真实数据集上的实验结果表明模糊C均值算法经该方法优化后,能快速收敛至全局最优解,在保证聚类速度的同时提高了聚类精度.
|
| 关 键 词： | 模糊C均值 粒子群算法 分集聚类 全局寻优 |
Novel Optimization Method for Fuzzy C-Means Algorithms |
| |
| CHU Hui-lin,ZHAO Bo,ZHANG Xing-ming.Novel Optimization Method for Fuzzy C-Means Algorithms[J].Journal of Information Engineering University,2011,12(3):347-351. |
| |
| Authors: | CHU Hui-lin ZHAO Bo ZHANG Xing-ming |
| |
| Affiliation: | CHU Hui-lin1,ZHAO Bo2,ZHANG Xing-ming3(National Digital Switching System Engineering & Technological R&D Center,Information Engineering University,Zhengzhou,450002,China) |
| |
| Abstract: | Considering fuzzy C-means clustering algorithms are sensitive to initialization and easy fallen to local minimum,a novel optimization method is proposed.In this method particle swarm is initialized through the result of subset clustering in order to get the near global optimal clustering center,under which clustering on full set is then done to avoid getting into local optimum.Experiments on real data set show that optimized fuzzy C-means algorithms quickly converge to global optimal solutions,guarantee the... |
| |
| Keywords: | fuzzy C-means particle swarm optimization subset clustering global optimization |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《信息工程大学学报》浏览原始摘要信息 |
|
点击此处可从《信息工程大学学报》下载全文 |
|
