An Improved K-means Clustering Algorithm Based on Meliorated Initial Centre

DOI

10.2991/aiie-16.2016.17How to use a DOI?

Keywords

clustering; outlier index; initial clustering center; k-means clustering

Abstract

The initial clustering center of traditional K-means clustering algorithm is selected at random that different initial clustering center will get different clustering results, which have great randomicity and poor stability. To improve the K-means clustering algorithm optimized by adopting local outlier index, we adopt a positive approach by calculating the local outlier index of all data samples. Then k local dense points with furthest mutual distance were selected as the initial clustering center. At last, in this paper we eliminate the effects of local outlier using in the improved algorithm. The experimental results showed that this enhanced algorithm could reduce the susceptibility by using K-means clustering algorithm to select the initial clustering center, and also short number of iterations. In brief, the K-means Clustering Algorithm Based on Meliorated Initial Centre obtains a more accurate clustering result.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Xiang Li
AU  - Zhenwei Wei
AU  - Lingling Li
PY  - 2016/11
DA  - 2016/11
TI  - An Improved K-means Clustering Algorithm Based on Meliorated Initial Centre
BT  - Proceedings of the 2016 2nd International Conference on Artificial Intelligence and Industrial Engineering (AIIE 2016)
PB  - Atlantis Press
SP  - 73
EP  - 76
SN  - 1951-6851
UR  - https://doi.org/10.2991/aiie-16.2016.17
DO  - 10.2991/aiie-16.2016.17
ID  - Li2016/11
ER  -