Open Access

Table 2.

Hyper-parameter setup for the DNN.

Hyper-parameter Setup
Number of hidden layers 4
Neurons per layer 64
Training epoch 10 000
Activation function ReLU
Batch size 8192
Optimiser Adam
Initial learning rate 10−4
Learning rate scheduler patience 300
Learning rate scheduler factor 0.8
Loss function MSE
Early stop 2000

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.