Forecasting petrol prices in India using neural networks models

Abstract

The work presented in this research paper constitutes a contribution to modeling and forecasting the monthly average petrol price per liter in Delhi using Feed Forward Neural Networks models. The model performance is on training and test samples measured using Mean Absolute Error (MAE), Mean Absolute and Percentage Error (MAPE,) and Root Mean Square Error (RMSE). The results show that feed forward neural networks (FFNN) models could be utilized to forecast the average petrol prices in Delhi city.