MALARIA DISEASE PREDICTION IN WEST AFRICA USING SELECTED MACHINE LEARNING TECHNIQUE

Abstract

Malaria, a life-threatening disease caused by Plasmodium parasite, remains a global health challenge with significant morbidity and modularity, particularly in sub-Saharan Africa. According to estimates from the World Health Organization (WHO), there were approximately 229 million clinical cases of malaria in 2019 and 409,000 deaths as a result (World Malaria Report, 2019). As a result of an increase in cases and fatalities, malaria is becoming a serious public health concern in West Africa. The focus of this study is to ensure machine learning can help people make a preliminary judgement about malaria according to their daily physical examination data and it can serve as a reference for doctors. The dataset was collected from Kaggle public repository and used to develop a predictive supervised machine learning models such as random forest, decision tree, k-nearest neighbor, artificial neural network and gradient boosting algorithms. Gradient boosting and Decision tree models were found to be the best performing model with an accuracy of 98.3% and 91.3% respectively. The evaluation metrics deployed for the study showed that RMSE (0.11), MAE (0.012), MSE (0.012), F1-score(0.80, 1.00). To further strengthen the evaluation method, confusion matrix produced TP 4 ,0, TN 1,76. The model will help health works, medical personnel and even the patients when diagnosing, to correctly predict Malaria among pateients suspected to have malaria