Evaluating Machine Learning Models for Rainfall Prediction: A Case Study of Nyando in Kenya

This paper presents a comprehensive evaluation of machine learning algorithms for rainfall prediction in the Nyando region. The study employs LSTM, XGBoost, Random Forest, and SVR algorithms, exploring both univariate and multivariate models to enhance the accuracy of predictions. Additionally, the paper examines three different outlier filtering methods and assesses their impact on the final prediction outcomes. The research endeavours to contribute valuable insights to the field of rainfall prediction and disaster management. By providing accurate and reliable rainfall predictions, this study aims to aid communities in the Nyando region and similar areas in their efforts to effectively mitigate the adverse impacts of extreme weather events.