Malaria outbreaks pose a significant health threat globally, especially in Sub-Saharan Africa and Nigeria, which accounts for over 25% of the global burden of the disease. According to the Malaria Indicator Survey of 2021, Sokoto State is one of the states with a high malaria burden (between 31% and 40% prevalence) in Nigeria. This study evaluated the potential of predictive analytics using machine learning to predict and prevent malaria outbreaks by exploring the relationship between malaria incidence and climatic data specifically temperature and rainfall.

This study employed a retrospective observational design to analyze historical malaria incidence and climatic data. The monthly malaria incidence data, spanning January 2015 to December 2022, were obtained from the Nigeria District Health Information System (DHIS) while the Climatic data, including temperature and rainfall, was collected from the National Aeronautics and Space Administration (NASA) website for the same period. The collected data was cleaned to remove any inconsistencies, missing values, or outliers. Three supervised machine learning algorithms were chosen; Support Vector Machine (SVM), Random Forest Classifier, and K-Nearest Neighbors (KNN). The models were trained using the preprocessed data, with hyperparameter tuning to optimize performance.

The results of the models revealed that SVM achieved the highest accuracy of 76% in forecasting malaria outbreaks, significantly outperforming both Random Forest and KNN, which achieved 59% accuracy. The trained models were evaluated using appropriate metrics such as accuracy, precision, recall, F1-score, and AUC-ROC.                            

This research demonstrates the effectiveness of machine learning, particularly SVM, in predicting malaria outbraks using climatic data in Sokoto State however, it is important to also consider other malaria-causing factors such as socioeconomic factors in subsequent studies. Overall, these models have the potential to equip people working to fight malaria with better information for preventing outbreaks and lessening their effects through early planning and actions.