Accession Number : AD0753715

Title :   Maximum Likelihood Estimation of a Translation Parameter of a Truncated Distribution,

Corporate Author : ILLINOIS UNIV URBANA DEPT OF MATHEMATICS

Personal Author(s) : Wolfowitz,J. ; Weiss,L.

Report Date : DEC 1972

Pagination or Media Count : 7

Abstract : In a very interesting paper Woodroofe obtained the limiting distribution of the maximum likelihood (m.l.) estimator of the translation parameter for a very interesting class of cases. This limiting distribution is normal with normalizing factor square root(n log n), so that one is in the domain of non-regular cases. Consequently, it is unknown whether the m.l. estimator is efficient and whether the practical statistician can use this estimator in applications. In the present paper the authors prove that it is efficient. The method of proof is to show that the m.l. estimator is equivalent to the maximum probability estimator, where R is any interval centered at the origin. The method is non-trivial and of wider applicability. (Author)

Descriptors :   (*STATISTICAL ANALYSIS, MATHEMATICAL PREDICTION), ASYMPTOTIC SERIES, RANDOM VARIABLES, PROBABILITY DENSITY FUNCTIONS, THEOREMS

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE