Two-Step MV-DeepONet: Probabilistic Operator Learning for Uncertainty Propagation Driven by Random Input Fields
Yupei Nie, Lei Wang, Jiasen Liu
Abstract
Forward uncertainty propagation in complex physical systems can induce structured covariance across field-valued outputs. For a probabilistic surrogate, the total predictive covariance comprises the covariance of conditional means across input realizations and the average conditional predictive covariance. Probabilistic DeepONet (Prob-DeepONet) provides lightweight uncertainty quantification by predicting pointwise Gaussian means and variances in a single forward pass, but its conditional predictive covariance is restricted to a diagonal form. To represent cross-location conditional dependence without explicitly parameterizing a full high-dimensional covariance matrix, we develop a two-step mean-variance DeepONet (two-step MV-DeepONet) through two principal modifications. First, two-step training is used to decouple output-basis learning from the input-to-coefficient mapping, together with basis orthogonalization and subspace rotation. Second, Gaussian probabilistic modeling is transferred from the high-dimensional physical output space to the low-dimensional rotated coefficient space. Mapping these probabilistic coefficients through the shared basis induces a generally non-diagonal conditional predictive covariance in the physical output space while retaining single-pass inference. A Frobenius-norm error decomposition and corresponding upper bound identify low-rank covariance compressibility, trunk-subspace approximation, finite-sample statistical error, and coefficient-space covariance estimation as the principal factors governing covariance recovery. Numerical experiments on three representative problems governed by partial differential equations (PDEs) and a hypersonic blunt-body aerothermal problem show improved generalization, more structured uncertainty bands, and accurate recovery of off-diagonal correlation patterns compared with Prob-DeepONet.
Create a lesson
Related papers
Graph lattice sums and graph zeta functions for long-range interacting quantum lattice models
Andreas Alexander Buchheit, Andreas Rupp
A numerical benchmark for fluid--structure--contact interaction
Daniele Corti, Jakub Fara, Miguel Angel Fernández et al.
Largest-dihedral-angle bisection algorithm does not preserve mesh regularity for tetrahedral partitions
Sergey Korotov, Jérôme Michaud
A Highly Scalable Quantized Tensor-Train FDTD Framework for the Simulation of Three-Dimensional Electromagnetic Scattering Problems
Daan Vanhaecke, Emile Vanderstraeten, Dries Vande Ginste
Pressure-robustness by commuting interpolation operators for Stokes discretizations with continuous pressures
Philip L. Lederer, Theresa Vock
A Reynolds-Semi-Robust, Globally Divergence-Free HDG Method for the Smagorinsky Model
Shuaijun Liu, Xiaoping Xie