Agricultural prices forecasting are the major concern for the policy makers as they directly affect the profitability of farming as an occupation. A hybrid model is considered to be an effective way to improve forecast accuracy. The hybrid model of the linear seasonal autoregressive moving average (SARIMA) and the nonlinear Artificial neural network (ANN) is proposed in this paper for estimating and forecasting the monthly wholesale price and arrival series of tomato crop. The goodness of fit of the model is measured using Akaike information criteria (AIC), root mean square error (RMSE), and mean absolute percentage error (MAPE), while post-sample forecast accuracy is measured using mean absolute error (MAE) and percent standard error of prediction (SEP). The study clearly shows that the hybrid (SARIMA-ANN) model is superior for forecasting the monthly wholesale prices and arrival series of tomato in the Gurugram market. The R (4.1.3) software is used for the analysis.

Price and arrival forecasting, MAE, SARIMA, SARIMA-ANN, and SEP

This study compared the modelling and forecasting performance of SARIMA and Hybrid (SARIMA-ANN) models using monthly wholesale price and arrival series of tomato crops in Gurugram market of Haryana. The goal of this study is short term forecast up to one year with different forecast horizons, such as 1, 3, 6, 9 and 12 months. SARIMA(1,1,2)(1,1,1)12 and SARIMA(2,1,2)(0,1,0)12 are the suitable models for capturing the linear pattern of price and arrival series, with the lowest AIC, RMSE, and MAPE values and significance parameter estimation. In comparison to the SARIMA and Hybrid models, the hybrid models provide better forecasting accuracy in terms of the lowest value of performance statistics such as MAE and SEP for 6, 9, and 12 months ahead forecast, whereas the SARIMA model performs better for 1 and 3 months ahead forecast.

Pushpa, Joginder Kumar and Vikram (2022). Hybrid Sarima-Ann Model for Forecasting Monthly Wholesale Price and Arrival Series of Tomato Crop. Biological Forum – An International Journal, 14(4a): 591-596.