A QR algorithm for the delta AR model assuming autocorrelation windowed data

A QR-type algorithm is developed to fit the delta autoregressive (DAR) model of R. Vijayan et al. to autocorrelation windowed sampled data. Vijayan et al. have developed Levinson-Durbin-type and Schur-type algorithms to compute the DAR model parameters when given a matrix Q _n, which takes the place of the conventional autocorrelation matrix R_n. They argue that the DAR model performs better than the conventional AR model for rapidly sampled data. There is not yet a theory on obtaining good estimates Q _n of from sampled data, contrasting with the well-developed theory for estimating R_n. The proposed QR-type algorithm overcomes this problem by computing the DAR model parameters without the need for estimating Q_n directly. The AR algorithm proposed is a simple modification of the classical QR algorithm for the classical AR model due to C.P. Rialan and L.L. Scharf (1988)