Estimation for finite parameter schemes

Author

Lii, K.S. ; Rosenblatt, M.

Author_Institution

California Univ., Riverside, CA, USA

fYear

1993

fDate

1993

Firstpage

129

Lastpage

130

Abstract

The object is to indicate the character of results for the approximate maximum likelihood estimation of parameters in nonminimum phase nonGaussian finite parameter schemes. The estimates are asymptotically normal under appropriate smoothness and positivity conditions on the probability density function of the generating independent random variables. The character of the asymptotic covariance matrix is indicated. In the truly nonminimum phase nonGaussian case one does not have consistency using the classical estimates using a Gaussian likelihood.

Keywords

approximation theory; maximum likelihood estimation; parameter estimation; Gaussian likelihood; approximate maximum likelihood estimation; asymptotic covariance matrix; asymptotically normal estimates; independent random variables; nonminimum phase nonGaussian finite parameter; parameter estimation; positivity conditions; probability density function; smoothness conditions; Covariance matrix; Deconvolution; Density functional theory; Equations; Maximum likelihood estimation; Parameter estimation; Phase estimation; Polynomials; Probability density function; Random variables;

fLanguage

English

Publisher

ieee

Conference_Titel

Higher-Order Statistics, 1993., IEEE Signal Processing Workshop on

Conference_Location

South Lake Tahoe, CA, USA

Print_ISBN

0-7803-1238-4

Type

conf

DOI

10.1109/HOST.1993.264582

Filename

264582