The Cholesky Decomposition Algorithm and ARMA Modeling

摘要

The history of the use of the Cholesky decomposition in the estimation and prediction of autoregressive moving average time series is reviewed and a modification of Ansley's (1979) estimation method is suggested. This modification is motivated by the results of Newton and Pagano (1981) in prediction theory and results in a significant reduction in the calculations required to evaluate the exact likelihood. It is also shown how the modified method can be used to obtain predictors and prediction variances. (Author)