CNN-Based Blood Group Prediction Using Fingerprint Images

This coursework explores blood group prediction using convolutional neural networks (CNN) and fingerprint patterns. It reviews studies on the connection between fingerprint patterns and blood, highlights CNN feasibility, and addresses gaps in manual methods. The methodology involves dataset preprocessing, optimized AlexNet, and Adam optimizer for training. Results achieved an 80% validation accuracy with robust performance and minimized overfitting. Future work includes testing simpler...