Figure 1: Comparison between predicted and reported data pressure drop for (a): train set; (b): validation set; (c): test set.