Autoregressive process parameters estimation under non-classical error model

Error in measuring time varying data setting is one important source of bias in estimating of time series modeling parameters. When the measurement error model is non-classic, this raises the question whether the different measurement error model strategy might differently affect the estimation of the time series modeling parameters. In this article, we investigate this in Autoregressive (AR) model parameters estimation under the non-classical measurement error model. We compare the parameters estimation of the AR model under the classical and non- classical error models. We perform analytically this on the AR model of order p. Further, we confirm this through simulation study specifically on the AR model of order 1. -- non-classical measurement error model ; autoregressive process ; parameter estimation