1: \begin{abstract}

2: Different variants of approximate inverse iteration like the 

3: locally optimal block preconditioned conjugate gradient method 

4: became in recent years increasingly popular for the solution 

5: of the large matrix eigenvalue problems arising from the 

6: discretization of selfadjoint elliptic partial differential 

7: equations, in particular for the calculation of the minimum 

8: eigenvalue. We extend in this little note the classical 

9: convergence theory of D'yakonov and Orekhov [Math. Notes 27 

10: (1980)] to the case of operators with an essential spectrum 

11: on infinite dimensional Hilbert spaces and allow for arbitrary, 

12: sufficiently small perturbations of the solutions of the 

13: equation that links the iterates. The note complements the 

14: much more elaborate convergence theory of Neymeyr and Knyazev 

15: and Neymeyr for the matrix case; see [Knyazev and Neymeyr, 

16: SIAM J. Matrix Anal. Appl. 31 (2009)] and the references 

17: therein. 

18: 

19: %

20: % Keywords:

21: %

22: % Eigenvalues, Eigenfunctions, Inverse Iteration 

23: %

24: %

25: % Mathematics Subject Classification (MSC2010):

26: %

27: 

28: \subclass{65N25 \and 65N15 \and 65N30}

29: 

30: %

31: % 65Nxx  Partial differential equations, boundary value problems

32: %

33: % 65N25  Eigenvalue problems

34: %

35: % 65N30  Finite elements, Rayleigh-Ritz and Galerkin methods

36: %

37: % 65N15  Error bounds

38: 

39: \end{abstract}

40: