On-Line K-PLANE Clustering Learning Algorithm for Sparse Comopnent Analysis

In this paper we propose a new algorithm for identifying mixing (basis) matrix A knowing only sensor (data) matrix X for linear model X = AS + E, under some weak or relaxed conditions, expressed in terms of sparsity of latent (hidden) components represented by the matrix S. We present a simple and efficient on-line algorithm for such identification and illustrate its performance by estimation of unknown matrix A and source signals S. The main feature of the proposed algorithm is its adaptivity to changing environment and robustness in respect to noise and outliers that do not satisfy sparseness conditions