A recursive algorithm for computing Cramer-Rao-type bounds on estimator covariance

A. Hero, J.A. Fessler
1994 IEEE Transactions on Information Theory  
We give a recursive algorithm to calculate submatriees of the Cramer-Rao (CR) matrix bound on the eovpriance of any unbiased estimator of a vector parameter 9. Our a≺&bm unmputea a scqpcnee of lcrwer bounds that oonvmges I P O~ to tBe CR b o d a#h exp@nenti.l speed of Ewvergence. The naueive PlgMithm uses an invertible "splitting matrix" to successively approximate the inverse Fisher information matrix. We present a statistical approach to selecting the splitting matrix based on a ' " n~e t e d
more » ... on a ' " n~e t e d a~i~t e d a t a~ tomulation similar to that of the well-known EM parameter estimation algorithm. As s c " t e Ustration we "@er iarrgc rceanstrpetlon f" projections for emission computed tomography.
doi:10.1109/18.335955 fatcat:nelx2rkz3zcdbgl545ifdmagb4