NSGA-PINN: A multi-objective optimization method for physics-informed neural network training

Published in Algorithms, Volume 16, Issue 4, Article 194 (2023)

NSGA-PINN introduces a Pareto-based optimization strategy to jointly minimize data and physics residuals in training physics-informed neural networks. This improves generalization and convergence in both forward and inverse problems.

Authors: B Lu, C Moya, G Lin
DOI: 10.3390/a16040194