Abstract
In this paper, two time series were used and applied in two methods, the first method is the method of the transfer function model and the second method is the method of artificial neural networks.The aim of This study to prepare a theoretical and practical study for predicting bivariate time series using the transfer function model as well as artificial neural network models to predict the annual production of rice crop in Iraq by analyzing a time series that spanned from 1971 to 2019 except for the Kurdistan region. The results were proven using the R program. The best transfer function model to predict the annual production of the rice crop in Iraq is the following model: -1.002B+ 0.078B^3) Zt+(1+0.188B-0.409B^2)/(1-0.498B-0.499B^2 ) a_tAs well as, by notethe MSE to the transformation function model and the MSE of the artificial neural network, the MSE is less when using the artificial neural networks model. Therefore, the best model for predicting the annual production of rice in Iraq is the artificial neural networks. .
