Effect of Activation Functions on Performance of Physics Informed Neural Networks: Beam Deflection Case Study: Supplementary Material

Results of model configurations and calculations have been presented as supplementary material for the study "Effect of Activation Functions on Performance of Physics Informed Neural Networks: Beam Deflection Case". The results of all 720 calculations are presented as supplementary material. The presented data includes model weights, loss history and deflection results history in raw format. The folder structure is ordered like “Results\ Boundary Condition Type\ Model Type\ Activation Function Type”. Loss history and deflection results are saved in python-pickle format and model weights are saved in pytorch pth format. Accuracy and rate of convergence values of each model are also included in the supplementary material in “Results\ Tables” folder as semicolon demitted text.