Comparative performance of ANN and ARIMA models in redgram price forecasting - Kalaburagi market

Abstract

The fluctuations in market arrivals largely contribute to price instability. Analysis of price overtime is important for formulating a sound agricultural policy. In view of this, the present study was undertaken by collecting monthly prices of redgram in Kalaburagi regulated Market of Karnataka for a period of 15 years (2002 to 2016). The purpose of this study was also to compare the forecasting performances of different time series methods for forecasting redgram prices. The various forms of Auto Regressive Integrated Moving Average (ARIMA) time series model and Artificial Neural Network (ANN) were employed to predict the future prices. On comparing the alternative models, it was observed that among ARIMA models Akaike Information criteria (4468.98) and Root Mean Square Error (291.74) were the least for ARIMA (3, 1, 2) model and ANN model with minimum RMSE value 244.01 and with highest R2= 98.17 per cent. The validity of the forecasted price values of redgram was checked and observed an accuracy of 95 to 98 per cent between actual and forecasted value. Therefore, price forecasting using ANN model was considered the most suitable for redgram in Kalaburagi Regulated market.