计算机工程与应用2011，Vol.47Issue(10)：35-37,157,4.DOI:10.3778/j.issn.1002-8331.2011.10.010
改进的核可能性C-均值聚类算法研究
Research of improved kernel possibilistie C-means clustering algorithm.
魏娜 1黄学宇 2高山3
作者信息
- 1. 西安空军工程大学,西安,710051
- 2. 西安空军工程大学导弹学院,西安,713800
- 3. 西安空军工程大学工程学院,西安,710038
- 折叠
摘要
Abstract
This paper proposes an improved algorithm,which is a Kernel Possibilistic C-Means(KPCM) clustering algorithm.The algorithm is a generalization of the PCM.By limiting the feasible region of the solutions of PCM and using the global optimization techniques(here the Simulated Annealing(SA) is used as an example) to solve the optimization problem, the improved algorithm inherits the advantages of robustness to noise and avoids the problem of generating coincident clusters,and can well find the global optimal solution, and speed up the convergence speed by decreasing the search range of the global optimization techniques.关键词
聚类算法/核可能性C-均值聚类/模拟退火Key words
clustering algorithm/Kernel Possibilistic C-means(KPCM) clustering/ Simulated Annealing(SA)分类
信息技术与安全科学引用本文复制引用
魏娜,黄学宇,高山..改进的核可能性C-均值聚类算法研究[J].计算机工程与应用,2011,47(10):35-37,157,4.
