Can physical information aid the generalization ability of Neural Networks for hydraulic modeling?

arXiv:2403.08589v1 Announce Type: new
Abstract: Application of Neural Networks to river hydraulics is fledgling, despite the field suffering from data scarcity, a challenge for machine learning techniques. Consequently, many purely data-driven Neural Networks proved to lack predictive capabilities. In this work, we propose to mitigate such problem by introducing physical information into the training phase. The idea is borrowed from Physics-Informed Neural Networks which have been recently proposed in other contexts. Physics-Informed Neural Networks embed physical information in the form …

abstract application arxiv capabilities challenge cs.lg data data-driven information machine machine learning machine learning techniques modeling networks neural networks physics.flu-dyn predictive type work