Estimation of the Parameters of a Modified Compound Poisson Distribution.
Abstract
This paper addresses the problem of estimation of the parameters of the Poisson sum of Gaussian random variables imbedded in a background of Gaussian noise when only realizations of the sum are observable. Cumulant matching, maximum likelihood, and an empirically orthogonalized characteristic function procedures are considered. The characteristic function and the maximum likelihood procedures produce similar results in a simulation study. However, the characteristic function procedure is computationally superior. Conditions under which all procedures are incapable of parameter estimation are discussed. Keywords: random variables, normal distribution.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1986
- Accession Number
- ADA178540
Entities
People
- A. S. Paulson
- J. L. Bryant