Parameter estimation techniques for autoregressive processes

Abstract

The main objective of this work is to find a more straightforward method for estimating the parameters of an equally spaced discrete autoregressive process by using maximum likelihood estimation (MLE) considering it is challenging to obtain the parameters of a nonlinear optimization procedure. The resulting estimated values are tested through simulation and then compared with those obtained using the previous MLE and Yule-Walker estimation. The achieved result yields slightly increased accuracy. Another problem we tackle is the Yule-Walker estimators for the continuous autoregressive models based on equally spaced discrete-time approximations. Again, these estimators are examined through simulation to demonstrate that the obtained result yields an accurate estimation.