Fig. 9: The training and validation of the mRMEBP framework.
From: mRMEBP: a unified framework for online detection of atrial fibrillation utilizing deep learning

a Illustration of the cross-entropy loss function \({\mathcal{L}}\). b Definition of local and global minimum points. c Parallelized algorithm for model diverse optimization. d Distribution of training outcome data. e Overview of statistical validation results.