Research map / Probabilistic and scientific modeling / Molecules and proteins
MMGP: a Mesh Morphing Gaussian Process-based machine learning method for regression of physical problems under nonparametrized geometrical variability
Fabien Casenave, Brian Staber, Xavier Roynard
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace best GNN baseline (GCNN) with MMGP | GCNN · Rotor37 m | relative RMSE: 4.4e-3 -> 5.0e-4 | Improved |
| MMGP vs best GNN baseline (MGN) on the hard plasticity scalar | MMGP · Tensile2d p_max | relative RMSE: 2.7e-1 -> 6.6e-1 | Mixed |
| replace best GNN baseline (MGN) with MMGP | MGN · AirfRANS C_D | relative RMSE: 4.9e-2 -> 3.3e-2 | Improved |
| replace best GNN baseline (MGN) with MMGP | MGN · AirfRANS C_L | relative RMSE: 2.4e-1 -> 8.0e-3 | Improved |
| replace best GNN baseline (GCNN) with MMGP | GCNN · Tensile2d v_max | relative RMSE: 4.4e-2 -> 5.0e-3 | Improved |
| MMGP vs MGN training time | MGN · Rotor37 | training time: 13 h 14 min -> 2 min 49 s | Improved |
| MMGP vs GCNN training time | GCNN · Rotor37 | training time: 24 h -> 2 min 49 s | Improved |
| MMGP vs GCNN training time | GCNN · Tensile2d | training time: 1 h 25 min -> 1 min 38 s | Improved |
| MMGP vs MGN training time | MGN · AirfRANS | training time: 5 h 00 min -> 5 min 47 s | Improved |
| MMGP vs GCNN Q2 on hardest field p | GCNN · Tensile2d p | Q2: 0.5691 -> 0.7785 | Improved |
| MMGP prediction intervals coverage at nominal 95% | MMGP · AirfRANS C_D | PICP: 93.5 | No clear effect |
| MMGP vs MGN on Rotor37 eta Q2 | MGN · Rotor37 eta | Q2: 0.5551 -> 0.9979 | Improved |
Closest papers by meaning
- MAVEN: A Mesh-Aware Volumetric Encoding Network for Simulating 3D Flexible Deformation (ICLR 2026 · similarity 0.86)
- Learning Mesh-Based Simulation with Graph Networks (ICLR 2021 · similarity 0.85)
- G-Adaptivity: optimised graph-based mesh relocation for finite element methods (ICML 2025 · similarity 0.85)
- PIG: Physics-Informed Gaussians as Adaptive Parametric Mesh Representations (ICLR 2025 · similarity 0.84)
- AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction (NeurIPS 2025 · similarity 0.84)
- Graph Neural PDE Solvers with Conservation and Similarity-Equivariance (ICML 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
