تحديد افضل مقدر خطي بطرق الانحدار الذاتي

Abstract

In this study, we use Levinson – Durbin Recurence which it used by Levinson in 1947, then its reformulation it later by Durbin in 1960, and compare it with Autoregressive Methods which include Least Squares, Yule-Walker equations and Burg methods. To find best linear predictor from the data which is represent the production of State Company for Cotton Industries. We concluded that Levinson – Durbin Recurrence is the best among to the other methods, according to minimum Mean Absolute Percentage Error (MAPE).