K-Subspace clustering and its application in sparse component analysis
Zhaoshui He, Andrzej S Cichocki · 2005
Abstract. The K-subspace clustering algorithm is established for sparse component analysis and overcome the difficulty that conventional SCA algorithms can not overcome. The conventional SCA algorithm can only perform single dominant SCA, can not perform multiple dominant SCA, but the proposed SCA algorithm based on K-subspace clustering can overcome this difficulty. 1