Towards an Efficient Prediction Model of Malaria Cases in Senegal

Abstract

One amongst the most deadly diseases in the world, Malaria remains a real flail in Sub-saharan Africa. In underdeveloped countries, e.g. Senegal, such a situation is acute due to the lack of high quality healthcare services and well-formed persons able to perform accurate diagnosis of diseases that patients suffer from. This requires to set up automated tools which will help medical actors in their decision making process. In this paper, we present first steps towards an efficient way to automatically diagnosis an occurence or not of Malaria based on patient signs and symptoms, and the outcome from the quick diagnosis test. Our prediction approach is built on the logistic regression function. First experiments on a real world patient dataset collected in Senegal, as well as a semi-synthetic dataset, show promising performance results regarding the effectiveness of the proposed approach.