The exact bayesian estimation technique of the error precision for MA (1)‎ process

Abstract EN

An important phase of time series analysis of moving average processes is to estimate the error precision.

This paper has three objectives.

The first one is to develop an exact Bayesian methodology to estimate the error precision of the first moving average process with unknown coefficient.

The second objective is to examine the numerical behavior of the proposed technique and its adequacy in solving the estimation problem.

The last objective is to compare the numerical efficiency of the proposed technique with the gamma approximation developed by Broemeling and Shaarawy (1988).

The numerical results support the adequacy of the proposed exact Bayesian technique and its superiority over the gamma approximation, particularly for small and moderate sample size.