Accuracy of Random Forest-Based Model for Malaria Parasite Prediction

Infectious diseases like Malaria have reportedly been the most prominent amongst the communicable diseases and have led to the death of approximately 435,000 annually in the world, while the majority of these fatalities occur in sub-Saharan Africa. Despite the deployment of heavy investment and strategy to mitigate or eradicate the malaria parasite, specifically in Sub-Saharan Africa, a high and upward trend of malaria cases is still being recorded. This study aims to examine the viability of the Random Forest Algorithm to predict the presence of the malaria parasite in patients accurately. The study presents a model based on the Random Forest Algorithm using MobileNetV2 as feature extraction. Meanwhile, the study considers some metrics, such as Precision, Recall and F1-Score, to further determine the performance of the model towards predicting the accuracy of malaria parasite in patients. The results confirm the accuracy of the Confusion Matrix classifier in predicting malaria parasite cases, while the model shows high accuracy and performance. The study contributes to the Health Informatics-Based Machine Learning domain towards predicting the Malaria parasite in patients.