|
|||||
|
|
| 基于改进微粒群算法的快速山峰聚类法 | |
| 引用本文： | 沈洪远,彭小奇,王俊年,刘健辰. 基于改进微粒群算法的快速山峰聚类法[J]. 系统工程学报, 2006, 21(3): 333-336 |
| 作者姓名： | 沈洪远 彭小奇 王俊年 刘健辰 |
| 作者单位： | 1. 中南大学能源与动力工程学院,湖南,长沙,410083;湖南科技大学信息与电气工程学院,湖南,湘潭,410201 2. 中南大学能源与动力工程学院,湖南,长沙,410083 3. 湖南科技大学信息与电气工程学院,湖南,湘潭,410201;中南大学信息科学与工程学院,湖南,长沙,411083 4. 湖南科技大学信息与电气工程学院,湖南,湘潭,410201 |
| 基金项目： | 国家自然科学基金;高等学校博士学科点专项科研项目 |
| 摘 要： | 依据样本密度使用高斯函数构造山峰函数时,削去对山峰函数贡献较小的大部分边缘,从而大大减少计算工作量;提出了一种改进的微粒群算法,使之具有多峰函数寻优能力,可以一次求出山峰函数的各个峰值,即基于改进微粒群算法的快速山峰聚类法,给出了算法的原理,步骤,快速山峰函数与常规山峰函数间的误差及计算工作量的比较.仿真结果表明,该算法计算简单快捷,可以一次求出所有的聚类中心,在满足精度要求的情况下,能够减少90%以上的计算工作量,有效地搜寻到数据样本空间的各个聚类中心,从而实现对数据样本的准确聚类. |
| 关 键 词： | 山峰聚类法 优化 微粒群算法 多峰函数寻优 高斯函数 |
| 文章编号： | 1000-5781（2006）03-0333-04 |
| 收稿时间： | 2005-03-01 |
| 修稿时间： | 2005-03-012005-09-12 |
Quick mountain clustering based on improved PSO algorithm |
|
| SHEN Hong-yuan,PENG Xiao-qi,WANG Jun-nian,LIU Jian-chen. Quick mountain clustering based on improved PSO algorithm[J]. Journal of Systems Engineering, 2006, 21(3): 333-336 | |
| Authors: | SHEN Hong-yuan PENG Xiao-qi WANG Jun-nian LIU Jian-chen |
| Affiliation: | 1. School of Energy and Power Engineering, Central South University, Changsha 410083, China; 2. School of Information and Electrical Engineering, Hunan University of Science and Technology, Xiangtan 411201, China; 3. School of Information Science and Engineering, Central South University, Changsha 410083, China |
| Abstract: | When a mountain function according to the density of the data samples is constructed by Gauss function,through cutting the most of the edge of the Gauss function the computational load is reduced mostly.A PSO algorithm is improved so that it is capable of multi-modal function optimization.Based on tahat a quick mountain clustering based on improved PSO(QMCBIPSO)algorithm is presented and its principle and steps are given.The difference of the quick mountain function and general mountain function and the comparison of their computational loads are indicated.The simulation experimental results show that the QMCBIPSO algorithm computes easily and can find all clustering centers of the data samples.For the same accuracy the computational load of the QMCBIPSO algorithm is 90% less than that of general algorithms.And it also can search efficiently and correctly every clustering center of the samples. |
| Keywords: | mountain clustering method optimization particle swarms optimization multi-modal function optimization Gauss function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |