Effort To Mitigate Malaria Via Early Detection Using Hybrid Machine Learning Architectures
Md. Sabbir Ahmed, Rafeed Rahman, Zarif Raiyan Arefeen, Auninda Alam, Marjan Tahreen · 2021
Malaria is a mosquito-borne disease spread by female Anopheles mosquitos that kills a large number of people every year, despite the fact that it has been around for over a century. The early identification of malaria patients is critical to minimizing the number of deaths caused by the disease. Manual malaria identification takes time and differs from expert to expert, thus an automated approach might save time while also being more efficient. This study proposes utilizing CNN models to identify malaria automatically from images of red blood cells. We utilized Decision Tree, KNN, SVM, AdaBoost, XGB classifier, and Random Forest for training after utilizing DenseNET-121 for feature extraction. The DenseNET-121+XGBClassifier model has the highest accuracy of 96.3% among the seven models suggested in this research.