Accession Number : ADA116943

Title :   Method of Moments Estimators for the Moving Average Coefficients in an ARMA (p,q) Process.

Descriptive Note : Technical rept.,

Corporate Author : SOUTHERN METHODIST UNIV DALLAS TEX DEPT OF STATISTICS

Personal Author(s) : Morton,M J ; Gray,H L

PDF Url : ADA116943

Report Date : May 1982

Pagination or Media Count : 17

Abstract : The standard procedure for computing the moving average method of moments parameter estimates for an ARMA process involves the numerical solution of a system of non-linear equations. In this paper it will be shown how the so-called inverse autocorrelations introduced by Cleveland can be utilized to compute the moving average parameter estimates as the solution to a system of linear equations, and that the method is faster and more accurate than solving the non-linear equations alluded to above. In addition, the procedure described below ensures that an invertible solution is obtained. (Author)

Descriptors :   *Maximum likelihood estimation, *Autocorrelation, *Moments, Solutions(General), Parameters, Computations, Methodology, Numerical methods and procedures, Linear differential equations, Inversion

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE