Malaria-Detection-using-Max-Voting-on-Fine-Tuned-Models

• Malaria is the pandemic disease. Major deaths are seen worldwide due to this disease.
• A blood sample is taken by the pathologist to detect the presence of parasite. This detection process requires a deep knowledge and experience. It challenges the accuracy of the report.
• Therefore, there is a need of some automated system that helps in decision making to detect it faster and with high precision. Many researches are done to provide effective automated solutions for detection and classification task.
• In this repo, three powerful pretrained deep neural networks are ensembled (bundled) together followed by max voting. The concept of fine tuning is used on these pretrained models. Overall, the proposed solution generated good results with an accuracy of 95.24%. Precision, recall, F1 and ROC scores are recorded as 94.57%, 96%,95.28% and 0.9524 respectively.