Deep UQ: learning deep neural network surrogate models for high dimensional uncertainty quantification
From MaRDI portal
Publication:2002273
DOI10.1016/J.JCP.2018.08.036zbMATH Open1419.68084arXiv1802.00850OpenAlexW2786232134MaRDI QIDQ2002273FDOQ2002273
Authors: Rohit K. Tripathy, I. Bilionis 
Publication date: 11 July 2019
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: State-of-the-art computer codes for simulating real physical systems are often characterized by a vast number of input parameters. Performing uncertainty quantification (UQ) tasks with Monte Carlo (MC) methods is almost always infeasible because of the need to perform hundreds of thousands or even millions of forward model evaluations in order to obtain convergent statistics. One, thus, tries to construct a cheap-to-evaluate surrogate model to replace the forward model solver. For systems with large numbers of input parameters, one has to deal with the curse of dimensionality - the exponential increase in the volume of the input space, as the number of parameters increases linearly. In this work, we demonstrate the use of deep neural networks (DNN) to construct surrogate models for numerical simulators. We parameterize the structure of the DNN in a manner that lends the DNN surrogate the interpretation of recovering a low dimensional nonlinear manifold. The model response is a parameterized nonlinear function of the low dimensional projections of the input. We think of this low dimensional manifold as a nonlinear generalization of the notion of the active subspace. Our approach is demonstrated with a problem on uncertainty propagation in a stochastic elliptic partial differential equation (SPDE) with uncertain diffusion coefficient. We deviate from traditional formulations of the SPDE problem by not imposing a specific covariance structure on the random diffusion coefficient. Instead, we attempt to solve a more challenging problem of learning a map between an arbitrary snapshot of the diffusion field and the response.
Full work available at URL: https://arxiv.org/abs/1802.00850
Recommendations
- Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
- Deep neural network surrogates for nonsmooth quantities of interest in shape uncertainty quantification
- Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
- Deep learning in high dimension: neural network expression rates for generalized polynomial chaos expansions in UQ
- An adaptive surrogate modeling based on deep neural networks for large-scale Bayesian inverse problems
Learning and adaptive systems in artificial intelligence (68T05) PDEs with randomness, stochastic partial differential equations (35R60) Numerical methods for partial differential equations, boundary value problems (65N99)
Cites Work
- Title not available (Why is that?)
- Adaptive subgradient methods for online learning and stochastic optimization
- Principal component analysis.
- Monte Carlo strategies in scientific computing.
- AlexNet
- Title not available (Why is that?)
- Gaussian processes for machine learning.
- Title not available (Why is that?)
- Regularization and Variable Selection Via the Elastic Net
- Gaussian processes with built-in dimensionality reduction: applications to high-dimensional uncertainty propagation
- Title not available (Why is that?)
- The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
- Multilayer feedforward networks are universal approximators
- Modeling uncertainty in flow simulations via generalized polynomial chaos.
- Predicting the output from a complex computer code when fast approximations are available
- Modeling uncertainty in steady state diffusion problems via generalized polynomial chaos
- Multi-output local Gaussian process regression: applications to uncertainty quantification
- Multi-output separable Gaussian process: towards an efficient, fully Bayesian paradigm for uncertainty quantification
- Uncertainty Quantification and Polynomial Chaos Techniques in Computational Fluid Dynamics
- Kernel principal component analysis for stochastic input model generation
- Large-Scale Machine Learning with Stochastic Gradient Descent
- Title not available (Why is that?)
- Discovering an active subspace in a single‐diode solar cell model
- Active subspace methods in theory and practice: applications to kriging surfaces
- Exploiting active subspaces to quantify uncertainty in the numerical simulation of the hyshot II scramjet
- Nonlinear methods of spectral analysis. 2nd corrected and updated edition
- Monte Carlo analysis of inverse problems
- Introduction to Uncertainty Quantification
- Multidimensional adaptive relevance vector machines for uncertainty quantification
- Inverse regression-based uncertainty quantification algorithms for high-dimensional models: theory and practice
- Uncertainty propagation using infinite mixture of gaussian processes and variational Bayesian inference
- Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
- Survey of Multifidelity Methods in Uncertainty Propagation, Inference, and Optimization
- Nonlinear information fusion algorithms for data-efficient multi-fidelity modelling
Cited In (only showing first 100 items - show all)
- Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data
- Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Solving the discretised neutron diffusion equations using neural networks
- Multifidelity multilevel Monte Carlo to accelerate approximate Bayesian parameter inference for partially observed stochastic processes
- Learning nonlocal constitutive models with neural networks
- Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons
- Train Like a (Var)Pro: Efficient Training of Neural Networks with Variable Projection
- Deep learning of free boundary and Stefan problems
- LaSDI: parametric latent space dynamics identification
- A bi-fidelity surrogate modeling approach for uncertainty propagation in three-dimensional hemodynamic simulations
- The Random Feature Model for Input-Output Maps between Banach Spaces
- GINNs: graph-informed neural networks for multiscale physics
- On generalized residual network for deep learning of unknown dynamical systems
- A physics-aware, probabilistic machine learning framework for coarse-graining high-dimensional systems in the small data regime
- A statistician teaches deep learning
- Image inversion and uncertainty quantification for constitutive laws of pattern formation
- Numerical solution of the parametric diffusion equation by deep neural networks
- Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
- Stein variational gradient descent with local approximations
- Mutual information for explainable deep learning of multiscale systems
- Understanding and Mitigating Gradient Flow Pathologies in Physics-Informed Neural Networks
- PPINN: parareal physics-informed neural network for time-dependent PDEs
- Efficient uncertainty quantification for dynamic subsurface flow with surrogate by theory-guided neural network
- Physics-informed machine learning models for predicting the progress of reactive-mixing
- Data-driven deep learning of partial differential equations in modal space
- Parametric dynamic mode decomposition for reduced order modeling
- Surrogate and inverse modeling for two-phase flow in porous media via theory-guided convolutional neural network
- Surrogate modeling for Bayesian inverse problems based on physics-informed neural networks
- A sample-efficient deep learning method for multivariate uncertainty qualification of acoustic-vibration interaction problems
- Data driven governing equations approximation using deep neural networks
- Deep ReLU neural network approximation in Bochner spaces and applications to parametric PDEs
- Enforcing Imprecise Constraints on Generative Adversarial Networks for Emulating Physical Systems
- An Adaptive Surrogate Modeling Based on Deep Neural Networks for Large-Scale Bayesian Inverse Problems
- Deep learning in high dimension: Neural network expression rates for generalized polynomial chaos expansions in UQ
- A higher-order singular value decomposition tensor emulator for spatiotemporal simulators
- Active learning Bayesian support vector regression model for global approximation
- Error estimates and physics informed augmentation of neural networks for thermally coupled incompressible Navier Stokes equations
- Machine learning for flux regression in discrete fracture networks
- Structured Bayesian Gaussian process latent variable model: applications to data-driven dimensionality reduction and high-dimensional inversion
- Bayesian neural networks for uncertainty quantification in data-driven materials modeling
- A survey of unsupervised learning methods for high-dimensional uncertainty quantification in black-box-type problems
- Learning constitutive relations from indirect observations using deep neural networks
- A Bayesian multiscale deep learning framework for flows in random media
- A novel sequential method to train physics informed neural networks for Allen Cahn and Cahn Hilliard equations
- An Acceleration Strategy for Randomize-Then-Optimize Sampling Via Deep Neural Networks
- Title not available (Why is that?)
- Coercing machine learning to output physically accurate results
- Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks
- Deep learning of vortex-induced vibrations
- ConvPDE-UQ: convolutional neural networks with quantified uncertainty for heterogeneous elliptic partial differential equations on varied domains
- Solving inverse problems using conditional invertible neural networks
- A deep learning driven pseudospectral PCE based FFT homogenization algorithm for complex microstructures
- Bayesian learning of orthogonal embeddings for multi-fidelity Gaussian processes
- Data-driven solutions and parameter discovery of the Sasa-Satsuma equation via the physics-informed neural networks method
- Extended dynamic mode decomposition for inhomogeneous problems
- Numerical weighted integration of functions having mixed smoothness
- Variational inference with NoFAS: normalizing flow with adaptive surrogate for computationally expensive models
- Stochastic modeling of inhomogeneities in the aortic wall and uncertainty quantification using a Bayesian encoder-decoder surrogate
- Deep capsule encoder–decoder network for surrogate modeling and uncertainty quantification
- Simulation of the 3D hyperelastic behavior of ventricular myocardium using a finite-element based neural-network approach
- A framework for strategic discovery of credible neural network surrogate models under uncertainty
- VI-DGP: a variational inference method with deep generative prior for solving high-dimensional inverse problems
- Learning to solve Bayesian inverse problems: an amortized variational inference approach using Gaussian and flow guides
- Wave-packet behaviors of the defocusing nonlinear Schrödinger equation based on the modified physics-informed neural networks
- slimTrain---A Stochastic Approximation Method for Training Separable Deep Neural Networks
- TDOR-MPINNs: multi-output physics-informed neural networks based on time differential order reduction for solving coupled Klein-Gordon-Zakharov systems
- Deep Neural Network Surrogates for Nonsmooth Quantities of Interest in Shape Uncertainty Quantification
- Random field of homogeneous and multi-material structures by the smoothed finite element method and Karhunen-Loève expansion
- Collocation approximation by deep neural ReLU networks for parametric and stochastic PDEs with lognormal inputs
- Operator learning using random features: a tool for scientific computing
- NeuralUQ: A Comprehensive Library for Uncertainty Quantification in Neural Differential Equations and Operators
- TNet: A Model-Constrained Tikhonov Network Approach for Inverse Problems
- Bayesian Deep Learning Framework for Uncertainty Quantification in Stochastic Partial Differential Equations
- Deep weighted Monte Carlo: a hybrid option pricing framework using neural networks
- Scalable conditional deep inverse Rosenblatt transports using tensor trains and gradient-based dimension reduction
- Level Set Learning with Pseudoreversible Neural Networks for Nonlinear Dimension Reduction in Function Approximation
- Neural and spectral operator surrogates: unified construction and expression rate bounds
- Bi-fidelity modeling of uncertain and partially unknown systems using DeepONets
- ReLU neural network Galerkin BEM
- Transformers for modeling physical systems
- Probabilistic partition of unity networks for high‐dimensional regression problems
- Deep Importance Sampling Using Tensor Trains with Application to a Priori and a Posteriori Rare Events
- Deep learning of biological models from data: applications to ODE models
- Dual order-reduced Gaussian process emulators (DORGP) for quantifying high-dimensional uncertain crack growth using limited and noisy data
- Quantification on the generalization performance of deep neural network with Tychonoff separation axioms
- An hp‐adaptive multi‐element stochastic collocation method for surrogate modeling with information re‐use
- Bi-fidelity variational auto-encoder for uncertainty quantification
- Polynomial chaos expansions on principal geodesic Grassmannian submanifolds for surrogate modeling and uncertainty quantification
- A deep neural network-based method for solving obstacle problems
- Sparse-grid sampling recovery and numerical integration of functions having mixed smoothness
- Accurate data-driven surrogates of dynamical systems for forward propagation of uncertainty
- Deep learning for thermal plasma simulation: solving 1-D arc model as an example
- Adaptive operator learning for infinite-dimensional Bayesian inverse problems
- Title not available (Why is that?)
- Bayesian neural networks for predicting uncertainty in full-field material response
- A review of recent advances in surrogate models for uncertainty quantification of high-dimensional engineering applications
- Uncertainty quantification for noisy inputs-outputs in physics-informed neural networks and neural operators
- From obstacle problems to neural insights: feedforward neural network modeling of ice thickness
- Data-driven modeling of partially observed biological systems
Uses Software
This page was built for publication: Deep UQ: learning deep neural network surrogate models for high dimensional uncertainty quantification
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2002273)
