The Modeling and Comparison of GMDH and RBF Artificial Neural Networks in Forecasting Consumption of Petroleum Products in the Agricultural Sector

Energy plays a significant role in today's developing societies. The role of energy demands to make decisions and policy with regard to its production, distribution, and supply. The vital importance of energy, especially fossil fuels, is a factor affecting agricultural production. This factor has a great influence on the production of agricultural products in Iran. The forecast of the consumption of oil products by the agricultural sector can help managers and planners to adopt sound management practices for their consumption. Presently, artificial neural networks are regarded as a powerful tool for the analysis and modeling of nonlinear relationships. The present study employed GMDH and RBF artificial neural networks to estimate the consumption of oil products by the agricultural sector. The underpinning parameters were selected to include the value added to the fixed price, rural population, agricultural land area, agricultural mechanization (tractor), and the consumption rate of oil products, electricity, price of oil products, and total energy use by the agricultural sector for the period of 1967-2017. The comparison of MSE, MAE, and MAPE for the GMDH and RBF models showed that the GMDH neural network was highly capable of modeling the energy consumption of the agricultural sector.