This research develops a machine learning model for drought prediction using the Random Forest algorithm, employing historical meteorological and soil data to deliver precise drought forecasts. Accurate drought prediction is essential for alleviating negative impacts on agriculture and water resources; however, conventional methods frequently lack precision. This study employs extensive data preprocessing, feature selection, and model training to develop a stable and interpretable predictive model. The algorithm, integrated with an interactive Streamlit application, allows stakeholders to submit data and receive real-time drought predictions. The evaluation criteria, such as accuracy, precision, and recall, demonstrate that the model successfully identifies the links between environmental variables and drought severity. The Random Forest model has robustness and interpretability, making it a significant asset for policymakers, agricultural planners, and researchers. This study also provides a user-friendly yet scientifically robust instrument for proactive drought management and highlights potential avenues for improved model precision and scalability.
Anika Pallapothu (Fri,) studied this question.