id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_vzc6tioriremjmtbvgl42jii34	Por-Shen Lai	Variance enhanced K-medoid clustering	2011	12	.pdf	application/pdf	8979	1127	75	This paper proposes new variance enhanced clustering methods to improve the popular K-medoid algorithm by adapting variance information in data clustering. model, groups similar images in clusters with various variances according to the distribution of image to cluster data points according to their pairwise similarity values. data point and its cluster center, the proposed method iteratively Using the proposed self-growing K-medoid algorithm, the number of required model (medoid) can be determined to let the dissimilarity between data point and closest cluster center be less In general, decision boundaries segment the feature space into polygon-shaped regions for each cluster in a data To select appropriate data points for the computation of variance along line segments between cluster centers, a polygonshaped data model, called Polygon descriptor (Lai & Fu, 2007, Data points are clustered using the K-medoid method proposed in Section 2.2.	./cache/work_vzc6tioriremjmtbvgl42jii34.pdf	./txt/work_vzc6tioriremjmtbvgl42jii34.txt