An Improved K-means Algorithm Based on Optimizing Initial Points

doi:10.7523/j.issn.2095-6134.2007.6.008

›› 2007, Vol. 24 ›› Issue (6): 771-777.DOI: 10.7523/j.issn.2095-6134.2007.6.008

• 论文 • Previous Articles Next Articles

An Improved K-means Algorithm Based on Optimizing Initial Points

QIN Yu, JING Ji-Wu, XIANG Ji, ZHANG Ai-Hua

The State Key Laboratory of Information Security(Graduate University of Chinese Academy of Sciences)

Received:1900-01-01 Revised:1900-01-01 Online:2007-11-15

Abstract

Abstract: K-means is an important clustering algorithm. It is widely used in Internet information processing technologies. Because the procedure terminates at a local optimum, K-means is sensitive to initial starting condition. An improved algorithm is proposed, which searches for the relative density parts of the database and then generates initial points based on them. The method can achieve higher clustering accuracies by well excluding the effects of edge points and outliers, as well as adapt to databases which have very skewed density distributions.

Key words: Clustering, K-means, Initial Points

CLC Number:

TP391

QIN Yu, JING Ji-Wu, XIANG Ji, ZHANG Ai-Hua. An Improved K-means Algorithm Based on Optimizing Initial Points[J]. , 2007, 24(6): 771-777.

[1]	YANG Suixin, GENG Xiurui, YANG Weitun, ZHAO Yongchao, LU Xiaojun. A method of hyperspectral remote sensing image classification based on spectral clustering [J]. , 2019, 36(2): 267-274.
[2]	LI Weiqian, QIU Ling. Joint clustering and resource allocation strategy for multicast device-to-device communication underlaying cellular networks [J]. , 2019, 36(1): 137-143.
[3]	XING Tao, HUANG Youhong, HU Qingrong, LI Jun, WANG Guanyong. SAR image segmentation based on dynamical K-means clustering algorithm [J]. , 2016, 33(5): 674-678.
[4]	CAO Laicheng, ZHAO Jianjun, CUI Xiang, LI Ke. Cyberspace device identification based on K-means with cosine distance measure [J]. , 2016, 33(4): 562-569.
[5]	GONG Xueshuang, YU Lijun, NIE Yueping, ZHU Jianfeng, PAN Yuqing . Spatial-pattern analysis of settlements in western region of Liaoning during pre-Qin period [J]. , 2016, 33(3): 373-379.
[6]	WU Wendi, CHENG Xijun, LIU Feng. CVaR portfolio model based on K-means clustering with the constraint of generalized entropy [J]. , 2016, 33(1): 31-36.
[7]	SUN Bo, GAO Sui-Xiang, LU Qing. Study on optimization of communication radii of nodes in hierarchically clustered wireless sensor networks [J]. , 2010, 27(6): 818-823.
[8]	WANG Qiu-Ming, GAO Hui-Ying, LIU Ke-Cheng. Software requirement analysis based on fuzzy clustering and grey correlation degree [J]. , 2010, 27(6): 859-863.
[9]	CAO Zheng, ZHU Ming. A fast and effective similar video search method [J]. , 2010, 27(3): 376-380.
[10]	MIAO Fu-You, XIONG Yan, WANG Xin-Xia. A 2-hop spanning tree based scalable clustered localization algorithm for wireless sensor networks [J]. , 2009, 26(6): 812-819.
[11]	XIA Lu-Ning, JING Ji-Wu. SA-DBSCAN:A self-adaptive density-based clustering algorithm [J]. , 2009, 26(4): 530-538.
[12]	WANG Jing, XIA Lu-Ning, JING Ji-Wu. Maximum density clustering algorithm [J]. , 2009, 26(4): 539-548.
[13]	SONG Jin-Liang, LUO Tie-Jian, CHEN Su, LIU Wei. A clustering based method to solve duplicate tasks problem [J]. , 2009, 26(1): 107-113.
[14]	LIU Wu-Hua, LUO Tie-Jian, WANG Wen-Jie. Quality Evaluation for Three Textual Document Clustering Algorithms [J]. , 2006, 23(5): 640-646.

An Improved K-means Algorithm Based on Optimizing Initial Points

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 14

Recommended Articles

Metrics

Comments