A 3D CNN surrogate predicts equivalent hydraulic conductivity tensors for discrete fracture-matrix systems with normalized RMSE below 0.22 and achieves over 100x GPU speedup.
Acta Numerica 24, 259–328
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
Multi-fidelity Monte Carlo using DSMC as high-fidelity and panel methods as low-fidelity control variates reduces relative RMSE in drag mean and second-moment estimates by factors of several when correlations are high.
MLMC and MLQMC with h- and p-refinement hierarchies achieve significant speedups over standard MC for UQ in cantilever beam problems, with MLQMC showing optimal cost scaling under certain conditions.
citing papers explorer
-
Convolutional Surrogate for 3D Discrete Fracture-Matrix Tensor Upscaling
A 3D CNN surrogate predicts equivalent hydraulic conductivity tensors for discrete fracture-matrix systems with normalized RMSE below 0.22 and achieves over 100x GPU speedup.
-
Multi-Fidelity Monte-Carlo Estimation of Satellite Drag in Very-Low-Earth Orbit
Multi-fidelity Monte Carlo using DSMC as high-fidelity and panel methods as low-fidelity control variates reduces relative RMSE in drag mean and second-moment estimates by factors of several when correlations are high.
-
h- and p-refined Multilevel Monte Carlo Methods for Uncertainty Quantification in Structural Engineering
MLMC and MLQMC with h- and p-refinement hierarchies achieve significant speedups over standard MC for UQ in cantilever beam problems, with MLQMC showing optimal cost scaling under certain conditions.