Skip to main content
. 2017 Dec 1;18(Suppl 13):478. doi: 10.1186/s12859-017-1878-3

Fig. 6.

Fig. 6

Loss of the model 4conv2pool4norm during training. The loss of the training set decreases rapidly, and we hold out a validation set for early stopping after 8 epochs of unimproved valid loss