A data-driven framework for sparsity-enhanced surrogates with arbitrary mutually dependent randomness

Abstract

The challenge of quantifying uncertainty propagation in real-world systems is rooted in the highdimensionality of the stochastic input and the frequent lack of explicit knowledge of its probability distribution. Traditional approaches show limitations for such problems, especially when the size of the training data is limited. To address these difficulties, we have developed a general framework of constructing surrogate models on spaces of stochastic input with arbitrary probability measure irrespective of the mutual dependencies between individual components of the random inputs and the analytical form. The present Data-driven Sparsity-enhancing Rotation for Arbitrary Randomness (DSRAR) framework includes a data-driven construction of multivariate polynomial basis for arbitrary mutually dependent probability measure and a sparsity enhancement rotation procedure.This sparsity enhancement method was initially proposed in our previous work [1] for Gaussian density distributions, which may not be feasible for non-Gaussian distributions due to the loss of orthogonality after the rotation. To remedy such difficulties, we developed a new data-driven approach to construct orthonormal polynomials for polynomials for arbitrary mutually dependent (amdP) randomness, ensuring the constructed basis maintains the orthogonality/near-orthogonality with respect to the density of the rotated random vector, where directly applying the regular polynomial chaos including arbitrary polynomial chaos (aPC) [2] shows limitations due to the assumption of the mutual independence between the components of the random inputs. The developed DSRAR framework leads to accurate recovery, with only limited training data, of a sparse representation of the target functions. The effectiveness of our method is demonstrated in challenging problems such as PDEs and realistic molecular systems within high-dimensional conformational space (O(10))where the underlying density is implicitly represented by a large collection of sample data, as well as systems with explicitly given non-Gaussian probabilistic measures. * huan.lei@pnnl.gov † The first two authors contributed equally ‡ nathan.baker@pnnl.gov arXiv:1804.08609v4 [math.NA]

Global sensitivity analysis: A Bayesian learning based polynomial chaos approach

“…However, it may happen in practical cases that the PDFs of the random variables are unknown. In that case, the polynomial basis functions can be constructed using the Gram-Schmidt orthogonalization, however, these are not always 'exact orthonormal' [39]. For that reason, a near optimal orthonormal polynomial has been developed in [39] and this can be used in conjunction with the proposed approach in this paper.…”

Data-driven molecular modeling with the generalized Langevin equation

“…Full atomic-detail molecular dynamics (MD) simulations are often prohibitively expensive due to the complexity and size of the systems under study. Model reduction based on surrogates [1][2][3][4][5][6] and projection operators [7,8] is a popular approach for reducing dimension and complexity in a wide range of computational science applications. One such model is the generalized Langevin equation (GLE) [7,8], which describes the system in terms of collective degrees of freedom and simulates dynamics in terms of coarse-grained collective variables (CVs).…”

“…In this work, we extend the data-driven parameterization approach [22] to construct a reduced model for the small molecule system of benzyl bromide (BnBr) in explicit water. We recently developed a data-driven approach [5] for uncertainty quantification of the equilibrium properties (e.g., solvation energy) with respect to the non-Gaussian conformation fluctuations using this solvated BnBr system. To quantify the non-equilibrium dynamics, the non-Markovian memory will need to be accurately constructed.…”

An efficient epistemic uncertainty quantification algorithm for a class of stochastic models: A post-processing and domain decomposition framework

“…Over the last two decades, the AUQ-PDE problem has been widely investigated using the Monte Carlo (MC), quasi-MC (QMC), and generalized polynomial chaos (gPC) techniques, see for example [6,7] (and references therein) for the MC, QMC, and gPC literature for forward AUQ computational models. In contrast, the literature on the EUQ-PDE forward problem is limited, see the recent work [8] and related EUQ references therein.…”