Iterative algorithms for computing the singular subspace of a matrix associated with its smallest singular values (Q1176528)

Three iterative methods --- inverse iteration, ordinary Chebyshev iteration and inverse Chebyshev iteration --- are discussed. The methods compute one singular vector (or a basis for a subspace of singular vectors belonging to several smallest singular values) instead of computing the complete singular value decomposition. The methods are compared by analyzing their convergence properties. Experimental results are presented in figures and tables.