1: \begin{abstract}

2: Convergence analysis of block iterative solvers for Hermitian eigenvalue problems

3: and the closely related research on properties of matrix-based signal filters

4: are challenging, and attract increasing attention due to their recent applications

5: in spectral data clustering and graph-based signal processing.

6: We combine majorization-based techniques pioneered for investigating

7: the Rayleigh-Ritz method in [SIAM J. Matrix Anal. Appl., 31 (2010), pp. 1521--1537]

8: with tools of classical analysis of the block power method by Rutishauser

9: [Numer. Math., 13 (1969), pp. 4--13] to derive convergence rate bounds

10: of an abstract block iteration, wherein tuples of tangents of principal angles

11: or relative errors of Ritz values are bounded using majorization in terms of

12: arranged partial sums and tuples of convergence factors. 

13: Our novel bounds are robust in presence of clusters of eigenvalues,

14: improve some previous results, and are applicable to

15: most known block iterative solvers and matrix-based filters,

16: e.g., to block power, Chebyshev, and Lanczos methods

17: combined with shift-and-invert approaches and polynomial filtering.

18: \end{abstract}

19: