Key points are not available for this paper at this time.
Abstract Potatoes are one of the widely consumed staple foods all over the world. The prices of potatoes were more unstable than other agricultural commodities due to factors such as perishability, production uncertainties, and seasonal fluctuations. These factors make it difficult for farmers to manage and predict production levels, resulting in supply and price fluctuations. Therefore, it is essential to develop predictive models that can accurately forecast the pricing of agricultural commodities like potatoes. The study attempted to explore the pattern of potato prices in major markets of northern India using different time series models. The empirical findings indicated positively skewed data distributed with a high instability index. In terms of forecasting accuracy, the EEMD-ANN model exhibited the best performance among the various time series techniques, generating the lowest MAPE values of 9.10%, 12.97%, and 4.27% for the Chandigarh, Delhi, and Shimla markets, respectively. Meanwhile, the EEMD-ARIMA model yielded the most accurate prediction results for the Dehradun market, with a MAPE value of 12.97%. The outcomes of this study offer significant insights to farmers, consumers, and government bodies for making informed decisions regarding the production, consumption, and distribution of potatoes. Moreover, the effectiveness of various time series models in handling complex agricultural price series was also investigated.
Shankar et al. (Mon,) studied this question.