Quantifying Uncertainty in Physics-Informed Neural Networks

Yip Jun Kai, Eduardo de Conto, Arvind Easwaran

Keywords: artificial intelligence, machine learning, physics-informed neural networks, uncertainty quantification, deep evidential regression

TL;DR: We integrate a state-of-the-art method to quantify aleatoric and epistemic uncertainties in physics-informed neural networks and observe that they can be captured effectively while maintaining predictive accuracy.

Confirmation Of Submission Requirements: I submit an abstract. It uses the template provided on the submission page and is no longer than 2 pages.

Submission Number: 208