Sparse Bayesian Learning Based on an Efficient Subset Selection

Liefeng Bo, Ling Wang, Licheng Jiao

Abstract: Based on rank-1 update, Sparse Bayesian Learning Algorithm (SBLA) is proposed. SBLA has the advantages of low complexity and high sparseness, being very suitable for large scale problems. Experiments on synthetic and benchmark data sets confirm the feasibility and validity of the proposed algorithm.