Phisics-informed neural networks, inverse problems, backward advection-dispersion equations, deep neural network training, importance sampling, parabolic equations
Geological carbon storage, deep learning, dimension reduction, principal component analysis, multilayer perceptron, 3D reconstruction model
Machine learning, streams and rivers, predictive solute transport, conservative solute transport
reduced-order models, inverse methods, time-dependent conditional Karhunen-Loeve expansions
redox flow battery, machine learning, energy storage, physics-constrained neural networks, electrochemical model
PINN method, parabolic PDEs, inverse PDEs, backward ADEs, DNN approximation
Model inversion, Gaussian process regression, conditional Karhunen-Lo\'{e}ve expansion, maximum a posteriori (MAP)
