Automatically finding clusters in normalized cuts

PATTERN RECOGNITION

卷:44

Article

Tepper, Mariano¹  Almansa, Andres²  Mejail, Marta^1,3 

[1] Univ Buenos Aires, FCEN, Dept Computac, RA-1053 Buenos Aires, DF, Argentina

[2] Telecom ParisTech, CNRS LTCI, Paris, France

[3] Univ Buenos Aires, Dept Comp Sci, Image Proc & Comp Vis Grp, RA-1053 Buenos Aires, DF, Argentina

关键词: Clustering;    Normalized cuts;    A contrario detection;   

DOI  :  10.1016/j.patcog.2011.01.003

来源: Elsevier

【 摘 要 】

Normalized Cuts is a state-of-the-art spectral method for clustering. By applying spectral techniques, the data becomes easier to cluster and then k-means is classically used. Unfortunately the number of clusters must be manually set and it is very sensitive to initialization. Moreover, k-means tends to split large clusters, to merge small clusters, and to favor convex-shaped clusters. In this work we present a new clustering method which is parameterless, independent from the original data dimensionality and from the shape of the clusters. It only takes into account inter-point distances and it has no random steps. The combination of the proposed method with normalized cuts proved successful in our experiments. (c) 2011 Elsevier Ltd. All rights reserved.

【 授权许可】

Free