On the Silhouette Plot in Cluster Analysis

• 발행기관: 한국자료분석학회
• 저자명: Daehak Kim
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.11 No.6, 2955~2963쪽, 전체 9쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2009.12.30

영문 초록

To find the appropriate number of clusters in a data set is very important issue in cluster analysis. In general, we do not know a real cluster structures, the number and composition of the groups for the given data. The silhouette plot is introduced and has been implemented to find the number of clusters with k-means clustering. Silhouette plot provides a very useful visible criterion to find the number of cluster in statistical cluster analysis. The silhouette plot is adapted to the well known real data sets to find the appropriate number of clusters.

목차

1. Introduction

2. K-means clustering and the number of clusters

3. Silhouettes

4. Numerical Data Analysis

5. Conclusion

References

키워드

cluster analysis hierarchical clustering k-means clustering number of clusters silhouettes