Time series forecasting of price for oilseed crops by combining ARIMA and ANN

Time series modelling and forecasting is a vibrant research field that had attracted the interest of scientific community in recent decades. Forecasts agricultural prices are proposed to be useful for farmers, governments, policy makers agribusiness industries. In this study, an effort made compare capabilities well-known linear Auto Regressive Integrated Moving Average (ARIMA) models, Delay Neural Network (TDNN) models Hybrid (ARIMA-TDNN) using data on monthly wholesale price four major oilseed crops India viz. groundnut, soybean, sesame rapeseed mustard from Jan-2001 Dec-2021. Finally, performance these evaluated compared by common criteria’s such as; Root Mean Square Error (RMSE), Absolute (MAE), Percentage (MAPE) percentage forecasts correct sign. Results showed lowest RMSE MAE were achieved hybrid model than ARIMA ANN all with exception MAPE which gave higher value sign highest others. Key findings revealed (ARIMA-ANN) outperformed each individual model, price.