A type of statistics first developed in the 19th Century could help improve our understanding of the spread of malaria, which very much remains a lethal infection in this century.
Researchers from Nigeria have employed a naïve Bayes as a probability classifier to help them predict whether or not new patients arriving with symptoms first actually have the parasitic disease and if they do what level of severity of infection and symptoms they are suffering. Such classification could help prioritise those patients who need urgent treatment.
The “framework” developed by the team has now been tested successfully on a sample dataset of some 700 records from a hospital in Yola, in Nigeria’s Adamawa State.
Aliyu, A., Prasad, R. and Fonkam, M. (2018) ‘A framework for predicting malaria using naïve Bayes classifier’, Int. J. Telemedicine and Clinical Practices, Vol. 3, No. 1, pp.78–93.