ABSTRACT
We observe X 1, …, Xn , where Xi = θ 1 Xi− 1 + Yi , where X 0 is defined as zero, and Y1,…, Yn are unobservable random variables, independent, each normal with mean zero and variance θ 2; θ 1 and θ 2 are both unknown and are to be estimated. By using maximum probability estimators, it is shown that the maximum likelihood estimators of the parameters have certain optimal properties.
