On Computing Multiple or Clustered Eigenvalues
Chang Wan Jeon, Jang Gyu Lee · Journal of Electrical Engineering and Information Science · 1999
The existing algorithms have presumed to apply themselves to matrices having distinct eigenvalues. However, many matrices have multiple eigenvalues and/or eigenvalues spaces so close that it is difficult to distinguish one from another. Two new algorithms for computing multiple and/or clustered eigenvalues are proposed in this paper. These algorithms are applicable to all kinds of Hermitian Toeplitz or real symmetric matrices unlike the existing algorithms. These algorithms are very effective for locating a small number of desired eigenvalues. Characteristics of them are investigated.