A Subspace Clustering Algorithm
Qiang Zhang · 2010
In this paper we present a new subspace clustering algorithm TGSCA for large dataset with noise. Experiments show that TGSCA can discover clusters both on entire space and subspace; the computation complexity is proximate linear with object's number, space dimension, and clusters' dimension respectively; it is not sensitive to noise; it can find both disjoint clusters or overlap clusters; it can find clusters of arbitrary shape; it is also able to find any number of clusters in any number of dimensions.