Enhancing wheat price forecasting with fuzzy time series: A case study from Birbhum District of West Bengal

Ensuring a stable and efficient market requires an in-depth analysis of commodity prices. Such analysis plays a crucial role in shaping strategic decisions and management policies by leveraging predictive and diagnostic tools to assess both current trends and future economic conditions of Birbhum district of West Bengal. This research focuses on applying fuzzy time series forecasting methods to predict wheat prices. A new methodology is introduced, utilizing fuzzy logic principles to improve prediction accuracy. To validate this approach, monthly wheat price data from January 2017 to December 2023 is analyzed. The study adopts the Chen method, where growth rate variations are used to define fuzzy sets within the universe of discourse. This technique, adapted from Chen’s model, employs the monthly price data as a foundation for forecasting future price levels. Findings indicate that this approach achieves greater precision, with a Mean Absolute Percentage Error (MAPE) of 7.66% and lower RMSE, outperforming other time series models. Furthermore, this study suggests refinements to the model to enhance forecasting reliability and enable more accurate one-step-ahead predictions. The findings of this research can support policymakers in developing data-driven pricing strategies, ensuring market stability, and optimizing supply chain decisions for wheat production.