Uncertainty quantification for noisy inputs–outputs in physics-informed neural networks and neural operators

Z Zou, X Meng, GE Karniadakis - Computer Methods in Applied Mechanics …, 2025 - Elsevier
Uncertainty quantification (UQ) in scientific machine learning (SciML) becomes increasingly
critical as neural networks (NNs) are being widely adopted in addressing complex problems …

Correcting model misspecification in physics-informed neural networks (PINNs)

Z Zou, X Meng, GE Karniadakis - Journal of Computational Physics, 2024 - Elsevier
Data-driven discovery of governing equations in computational science has emerged as a
new paradigm for obtaining accurate physical models and as a possible alternative to …

Leveraging viscous Hamilton–Jacobi PDEs for uncertainty quantification in scientific machine learning

Z Zou, T Meng, P Chen, J Darbon… - SIAM/ASA Journal on …, 2024 - SIAM
Uncertainty quantification (UQ) in scientific machine learning (SciML) combines the powerful
predictive power of SciML with methods for quantifying the reliability of the learned models …

NeuroSEM: A hybrid framework for simulating multiphysics problems by coupling PINNs and spectral elements

K Shukla, Z Zou, CH Chan, A Pandey, Z Wang… - Computer Methods in …, 2025 - Elsevier
Multiphysics problems that are characterized by complex interactions among fluid dynamics,
heat transfer, structural mechanics, and electromagnetics, are inherently challenging due to …

A comprehensive and FAIR comparison between MLP and KAN representations for differential equations and operator networks

K Shukla, JD Toscano, Z Wang, Z Zou… - arxiv preprint arxiv …, 2024 - arxiv.org
Kolmogorov-Arnold Networks (KANs) were recently introduced as an alternative
representation model to MLP. Herein, we employ KANs to construct physics-informed …

Large scale scattering using fast solvers based on neural operators

Z Zou, A Kahana, E Zhang, E Turkel, R Ranade… - arxiv preprint arxiv …, 2024 - arxiv.org
We extend a recently proposed machine-learning-based iterative solver, ie the hybrid
iterative transferable solver (HINTS), to solve the scattering problem described by the …

HJ-sampler: A Bayesian sampler for inverse problems of a stochastic process by leveraging Hamilton-Jacobi PDEs and score-based generative models

T Meng, Z Zou, J Darbon, GE Karniadakis - arxiv preprint arxiv …, 2024 - arxiv.org
The interplay between stochastic processes and optimal control has been extensively
explored in the literature. With the recent surge in the use of diffusion models, stochastic …