Open Access
Table A.1.
Parameter settings for the Artificial Neural Network (ANN) model implementation.
| Model parameters | Settings |
|---|---|
| Total hidden layers | 4 |
| Hidden layer activation function | Rectified linear unit |
| Output layer activation function | Softmax |
| Weight regularization | L2 with a penalty of 10−2 |
| Training batch size | 64 |
| Loss function | Weighted Binary Cross-Entropy (weights: wno−flare = 1, wflare = 100) |
| Optimization method | Adagrad (parameters: learning rate = 5 × 10−4) |
| Number of Epochs | 50 |
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.