| 基于遗传算法的K-means初始化EM算法及聚类应用 |
| |
| 引用本文： | 山拜·达拉拜,曹红丽,尤努斯·艾沙. 基于遗传算法的K-means初始化EM算法及聚类应用[J]. 现代电子技术, 2010, 33(15): 102-103,106 |
| |
| 作者姓名： | 山拜·达拉拜 曹红丽 尤努斯·艾沙 |
| |
| 作者单位： | 新疆大学,信息科学与工程学院,新疆,乌鲁木齐,830046 |
| |
| 基金项目： | 国家自然科学基金资助项目 |
| |
| 摘 要： | 混合高斯模型能够有效地拟合概率密度函数,常用的混合高斯概率密度模型参数估计方法是EM迭代算法,这种算法的缺点是估计精度过分依赖于初始值,而且不能估计模型阶数。基于遗传算法的K-means初始化EM算法可以同时估计模型阶数和参数。试验结果表明,该算法具有更好的聚类效果。
|
| 关 键 词： | 混合高斯模型 遗传算法 K-means 聚类应用 |
K-means Initialization EM Algorithm and Its Clustering Application Based on Genetic Algorithm |
| |
| SENBAI Dalabaev,CAO Hong-li,YOUNUSI Aisha. K-means Initialization EM Algorithm and Its Clustering Application Based on Genetic Algorithm[J]. Modern Electronic Technique, 2010, 33(15): 102-103,106 |
| |
| Authors: | SENBAI Dalabaev CAO Hong-li YOUNUSI Aisha |
| |
| Affiliation: | SENBAI Dalabaev,CAO Hong-li,YOUNUSI Aisha(School of Information Science ＆ Engineering,Xinjiang University,Urumqi 830046,China) |
| |
| Abstract: | The probability density function can be efficiently fitted by Gaussian mixture model.EM iteration algorithm is one of popular algorithms for parameters estimation of Gaussian mixture probability density model.However,this method de-pends on initial parameters highly and can not estimate the orders of models.K-means initialization EM algorithm based on the genetic algorithms can estimate the orders and parameters of models.The simulation results indicate that this method has a good clustering ability. |
| |
| Keywords: | K-means |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
