Skip to main content
Fig. 3 | BioData Mining

Fig. 3

From: Interpretable recurrent neural network models for dynamic prediction of the extubation failure risk in patients with invasive mechanical ventilation in the intensive care unit

Fig. 3

Instance of RNN training and early stopping. In each epoch, the model iteratively tuned its parameters by analyzing batches of sequences until reading through all sequences in the training set. As epochs increased, the cross-entropy loss in the training fold decreased gradually, whereas the loss in the validation fold firstly decreased and then increased. Early stopping was triggered if the loss in validation fold began to rise, in order to avoid over-fitting of the model

Back to article page

BioData Mining

ISSN: 1756-0381