Showing cs.LGShow all
2 papers · 1 filter
cs.LG2021
Effects of boundary conditions in fully convolutional networks for learning spatio-temporal dynamics
Antonio Alguacil, Wagner Gonçalves Pinto, Michael Bauerheim +2
Accurate modeling of boundary conditions is crucial in computational physics. The ever increasing use of neural networks as surrogates for physics-related problems calls for an imp…
cs.LG2021
On the reproducibility of fully convolutional neural networks for modeling time-space evolving physical systems
Wagner Gonçalves Pinto, Antonio Alguacil, Michaël Bauerheim
Reproducibility of a deep-learning fully convolutional neural network is evaluated by training several times the same network on identical conditions (database, hyperparameters, ha…