Accession Number : AD0691214

Title :   ESTIMATION OF PARAMETERS WITH INCOMPLETE DATA.

Descriptive Note : Final rept. on Themis optimization research program,

Corporate Author : TEXAS A AND M UNIV COLLEGE STATION INST OF STATISTICS

Personal Author(s) : Hocking,R. R. ; Smith,W. B. ; Waldron,B. R. ; Oxspring,H. H.

Report Date : JUL 1969

Pagination or Media Count : 70

Abstract : A general approach is described for the problem of estimating parameters in a multivariate distribution with incomplete or fragmentary data. Special attention is given to the case of multivariate normal populations but an example with the multinomial distribution is given to indicate the generality of the method. The approach taken in this paper is also applicable to the problem of optimally combining information from different sources when the different sources may provide only fragmentary information.

Descriptors :   (*MULTIVARIATE ANALYSIS, STATISTICAL DISTRIBUTIONS), (*STATISTICAL DISTRIBUTIONS, DECISION THEORY), STATISTICAL DATA, REGRESSION ANALYSIS, POPULATION(MATHEMATICS), THEOREMS

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE