TOWARDS CLASSIFICATION OF SYMPTOMATIC AND CLIMATIC BASED MALARIA PARASITE-COUNT

Abstract

Good health is one of the most important things in life. Several diseases affect the proper functioning of human’s health; one of such common disease is malaria. Malaria is a leading public health problem in the developing countries and Nigeria, leading to high morbidity and mortality and huge cost for diagnosis, treatment and control. The insurgency of malaria diseases has pushed the need to develop computational approaches for predicting the severity of malaria diseases based on symptoms and climatic factors. The prediction of the occurrence of malaria disease and its outbreak will be helpful to take appropriate precaution measures. The existing predicting models examine binary cases of malaria, prone to error, and suffer from overfitting due to large number of parameters to fix. This paper proposes a Support Vector Machine (SVM) with best activation function to determine the rate of malaria transmission. This paper aims to study the components of learning parameters in multiclass Support Vector Machine (SVM), study optimal separation hyperplane, review SVM classification and generate SVM malaria model. Monthly averages of rainfall, temperature, relative humidity and malaria serves as the input variables. Apart from classification, the future work will be based on using SVM for other machine learining technique functions such as pattern recognition, regression analysis and feature selection.