A Graphical User Interface Based on Logistic Regression Approach for Malarial Detection

Abstract

Malaria (a mosquito-infected disease) is one of the deadliest communicable diseases in the world. The disease causes a significant global health challenge. According to the World Health Organisation (WHO), millions of deaths occur every year worldwide. The mortality rate poses a challenge to authority and management. Over the years, mathematical and machine learning (ML)-based techniques have been developed to mitigate the scenario. In this study, ML-based prediction techniques are investigated to predict the presence of malaria in individuals. More specifically, three ML-based techniques—Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF)—are employed to differentiate their prediction performance (namely, classification accuracy, precision, recall, and F-score) over a created database (D) consisting of 350 records. Among the adopted techniques, the LR technique shows overall better performance over the test data chosen from D. A graphical user interface (GUI) based on LR is also developed to detect the presence or absence of malaria in any individual. The time spent by the GUI to report the absence or presence of the disease is definitely less than the time spent by malaria experts.