Malaria is a life-threatening disease, caused by a parasite that is transmitted to humans through the bite of an infected female Anopheles mosquito. In 2019, there were an estimated 229 million cases of malaria worldwide and the death toll reached 409,000. The area most frequently affected by malaria, according to WHO, is the African region. Malaria can be detected beforehand by using the information inpatient data and applying machine learning techniques. This study aims to detect and classify severe malaria based on the history of examining patient data using the Support Vector Machine (SVM) method with a normalization technique using min-max on the dataset and a cross-validation technique with several experiments on the K value of the results. This study also compares the Support Vector Machine method with Naïve Bayes (NB) where the accuracy of the SVM model is superior to Nave Bayes with an average accuracy gap of 25%. The accuracy generated by the application of the proposed method is 92.3%.
Copyrights © 2021