Abstract: Malaria is a disease caused by protozoan parasites of the genus Plasmodium that are transmitted through the bites of infected female mosquitoes. It mainly infects the red blood cells in the human body. Initially malaria may cause no symptoms but gradually it may throw up symptoms such as vomiting, headaches and eventually may result in coma. Some current malaria detection techniques include manual microscopic examination and Rapid Diagnostic Test. These approaches are vulnerable to human errors. Early detection of malaria can help in reducing the fatality rates and examination of features in blood cells may aid in vaccine development. This research work focuses on identifying the type of plasmodium using suitable machine learning model that would classify the type of plasmodium with greater accuracy. Binary classification using machine learning algorithms namely Convolutional Neural Network (CNN), Support Vector Machines(SVM), Random Forest(RF), Decision Tree(DT) and ensemble technique are implemented to predict the model with higher accuracy. A comparative analysis is performed to find the model with greater accuracy. The comparative analysis reveal that the CNN model has outperformed the other models with highest accuracy for each of the species classification.

Keywords: Histogram Oriented Gradients, Gabor, Kaze, Support Vector Machines, Decision Tree, Random Forest, Ensemble model, Convolutional Neural Network.

| DOI: 10.17148/IARJSET.2021.8517