
Accession Number : AD0697078
Title : COVARIANCE MATRIX ESTIMATION IN LINEAR MODELS,
Corporate Author : PAN AMERICAN WORLD AIRWAYS INC PATRICK AFB FLA
Personal Author(s) : Chew,Victor
Report Date : NOV 1967
Pagination or Media Count : 23
Abstract : In regression analysis with heteroscedastic and/or correlated errors, the usual assumption is that the covariance matrix of the errors is completely specified, except perhaps for a scalar multiplier. This condition is relaxed in this paper by assuming only that the covariance matrix has a certain pattern; for example, that the covariance matrix is diagonal or partitionable into a diagonal matrix of submatrices. The method used for estimating the covariance matrix is the standard procedure of equating certain quadratic forms of the observations (in this case, squares and products of residuals from regression) to their expectations, and solving for the unknown variances and covariances. A numerical example illustrates the method. (Author)
Descriptors : (*REGRESSION ANALYSIS, MATRICES(MATHEMATICS)), ANALYSIS OF VARIANCE, CORRELATION TECHNIQUES, DECISION THEORY, LEAST SQUARES METHOD, GUIDED MISSILE TRACKING SYSTEMS
Subject Categories : Statistics and Probability
Distribution Statement : APPROVED FOR PUBLIC RELEASE