-

Nowadays in commercial- economical subjects, prediction is known as one of the main branch of science and develops day after day. Managers of different economics and commercial sectors, because of many effective variables, prefer to have a mechanism which can help them in decision making. Agriculture as a sector which produces strategic products and provides needed foods for increasing population in society, have a main effect on significant of economic, social and political decision making. Analysis of growth and inflation rate variables in Agriculture sector and being aware of there future trend and also knowing structural data generating models can help politicians and planners in suitable decision making.
In this study we have used holt- winters exponential smoothing and ARIMA models. Monthly data for inflation rate variable during 1363-1383 were gathered. In order to make a comparison between linear and non- linear models prediction accuracy, we construct an artificial neural network according to regression model variables and ARIMA model. Result showed that holt-winters exponential smoothing has the most prediction accuracy of all. In our suggested model, agricultural sector average growth rate for the forth development program and average inflation rate in agricultural sector are predicted about 7% respectively.
JEL Classification: H12