Subspace iterative methods for eigenvalue problems

This paper presents novel perturbation bounds for generalized symmetric pos itive definite eigenvalue problems. The bounds provide the insights for an observed computational phenomenon that is not easily explained by the exist ing bounds developed previously. Using the new bounds, we provide an analys is of a subspace Newton type procedure for computing a few extreme eigenpai rs for generalized symmetric positive definite systems. A preconditioned ve rsion of this subspace iterative method is also studied. (C) 1999 Elsevier Science Inc. All rights reserved.