A machine learning approach for efficient uncertainty quantification using multiscale methods
From MaRDI portal
Publication:1700748
Abstract: Several multiscale methods account for sub-grid scale features using coarse scale basis functions. For example, in the Multiscale Finite Volume method the coarse scale basis functions are obtained by solving a set of local problems over dual-grid cells. We introduce a data-driven approach for the estimation of these coarse scale basis functions. Specifically, we employ a neural network predictor fitted using a set of solution samples from which it learns to generate subsequent basis functions at a lower computational cost than solving the local problems. The computational advantage of this approach is realized for uncertainty quantification tasks where a large number of realizations has to be evaluated. We attribute the ability to learn these basis functions to the modularity of the local problems and the redundancy of the permeability patches between samples. The proposed method is evaluated on elliptic problems yielding very promising results.
Recommendations
- Data-driven acceleration of multiscale methods for uncertainty quantification: application in transient multiphase flow in porous media
- Multiscale finite element methods for stochastic porous media flow equations and application to uncertainty quantification
- Machine-learning-based modeling of coarse-scale error, with application to uncertainty quantification
- A data-driven approach for multiscale elliptic PDEs with random coefficients based on intrinsic dimension reduction
- Dynamic data-driven Bayesian GMsFEM
Cites work
- A multiscale finite element method for elliptic problems in composite materials and porous media
- A multiscale restriction-smoothed basis method for high contrast porous media represented on unstructured grids
- Adaptive subgradient methods for online learning and stochastic optimization
- Approximation by superpositions of a sigmoidal function
- Deep learning
- scientific article; zbMATH DE number 6378127 (Why is no real title available?)
- Hyperband: a novel bandit-based approach to hyperparameter optimization
- Learning representations by back-propagating errors
- Mixed Multiscale Finite Element Methods for Stochastic Porous Media Flows
- Multi-scale finite-volume method for elliptic problems in subsurface flow simulation.
- Multilayer feedforward networks are universal approximators
- Multiscale finite-volume method for compressible multiphase flow in porous media
- Multiscale finite-volume method for density-driven flow in porous media
- Nonparametric guidance of autoencoder representations using label information
Cited in
(29)- Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
- A heterogeneous multiscale MPFA method for single-phase flows in porous media with inertial effects
- Machine-learning-based modeling of coarse-scale error, with application to uncertainty quantification
- Machine learning for flux regression in discrete fracture networks
- A Bayesian multiscale deep learning framework for flows in random media
- Multiscale model reduction for stochastic elasticity problems using ensemble variable-separated method
- Data-driven deep learning of partial differential equations in modal space
- On generalized residual network for deep learning of unknown dynamical systems
- Extended dynamic mode decomposition for inhomogeneous problems
- Data-driven polynomial chaos expansion for machine learning regression
- Data driven governing equations approximation using deep neural networks
- Regression-based sparse polynomial chaos for uncertainty quantification of subsurface flow models
- Machine learning for accelerating macroscopic parameters prediction for poroelasticity problem in stochastic media
- Data-driven acceleration of multiscale methods for uncertainty quantification: application in transient multiphase flow in porous media
- A multiscale neural network based on hierarchical nested bases
- Dynamic data-driven Bayesian GMsFEM
- Deep learning of biological models from data: applications to ODE models
- A survey of unsupervised learning methods for high-dimensional uncertainty quantification in black-box-type problems
- LaSDI: parametric latent space dynamics identification
- Interfacing finite elements with deep neural operators for fast multiscale modeling of mechanics problems
- Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons
- Multiscale elliptic PDE upscaling and function approximation via subsampled data
- A low-rank approximated multiscale method for PDEs with random coefficients
- Neural Network Approximation of Coarse-Scale Surrogates in Numerical Homogenization
- A machine learning-based probabilistic computational framework for uncertainty quantification of actuation of clustered tensegrity structures
- Deep-learning-based upscaling method for geologic models via theory-guided convolutional neural network
- Data-driven modeling of partially observed biological systems
- Adaptive and parallel multiscale framework for modeling cohesive failure in engineering scale systems
- A hybrid approach combining DNS and RANS simulations to quantify uncertainties in turbulence modelling
This page was built for publication: A machine learning approach for efficient uncertainty quantification using multiscale methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1700748)
