Performance of 1DCNN with fine-tuned hyperparameters. The figure illustrates the performance of 1DCNN models trained with a batch size of 256, a learning rate of 10⁻⁴, and a kernel size of 9 on EPG signal data. The left panel shows the loss (solid lines) and accuracy (dotted lines) curves during the training (blue) and validation (red) phases. These curves provide insights into the model’s learning progress and ability to generalize to unseen data. The middle panel presents evaluation metrics for both tasks at epoch 100. The right panel shows the confusion matrix for the classification task, revealing the per-class accuracy and any misclassifications between waveform categories.

<p>Performance of 1DCNN with fine-tuned hyperparameters. The figure illustrates the performance of 1DCNN models trained with a batch size of 256, a learning rate of 10⁻⁴, and a kernel size of 9 on EPG signal data. The left panel shows the loss (solid lines) and accuracy...

Full description