Predictive Analytics for Agricultural Futures: Modelling and Forecasting Net Exports of Commodities

Agricultural futures markets play a pivotal role in global trade, enabling producers, consumers, and investors to manage risks and capitalize on market opportunities. Accurate prediction of future exports is crucial for informed decision-making in this domain. In this research, we present a comparative analysis of several forecasting algorithms applied to “Agricultural Commodities Futures Data” sourced from Kaggle. The aim is to predict net exports for the next five years based on the selected commodity. We employ a range of models including Integrated Autoregressive Moving Average (ARIMA), Seasonal Autoregressive Integrate Moving-Average with Exogenous Regressors (SARIMAX), Long Short-Term Memory (LSTM) Least Absolute Shrinkage and Selection Operator (LASSO), Random Forest, and Local regression using LOESS combined with seasonal-trend decomposition. Through a Streamlit application, users can interactively input commodity names and receive forecasts generated by the various models.