A Prediction Model for Malaria using an Ensemble of Machine Learning & Hydrological Drought Indices

Опубликовано 20 сентября 2022, 15:12

Literature is awash with evidence that points to climate change as one of the reasons for the upward increase in infectious disease outbreaks around the globe. With the advancement in Artificial Intelligence (AI) and big climate data, accurate analysis of medical and climate data allows prediction and early detection of infectious diseases associated with climate change. The effectiveness of AI and drought indices in predicting oncoming droughts has been demonstrated elsewhere in research, this has however not been explored in the prediction of infectious diseases. We present a model that uses machine learning to predict the outbreak of infectious diseases using two drought indices and historical incidents of malaria in Limpopo (South Africa). Having achieved up to 99% prediction accuracies, we demonstrate that such a model equips stakeholders with a new perspective when it comes to prediction of infectious diseases’ outbreaks that are associated with extreme climate variations.

Professor Muthoni Masinde is an established computer science researcher under South Africa’s National Research Foundation (NRF) category C2 rating. As well as working as a Professor and the Head of the IT Department at the Central University of Technology, Free State (South Africa), she is the Founder and CEO of ITIKI - acronym for Information Technology and Indigenous Knowledge with Intelligence. Her accomplishments are based on the socio-economic impacts that the application of her research has had on small-scale farmers. ITIKI is the vehicle for her research; she developed ITIKI by integrating both the indigenous and scientific information on weather. This integration was made possible through the application of artificial neural networks, wireless sensor networks and mobile phone applications.

Learn more about MARI: microsoft.com/en-us/research/g...