Abstract:Abstract The vast majority of stochastic optimization problems require the approximation of the underlying probability measure, e.g., by sampling or using observations. It is therefore crucial to understand the dependence of the optimal value and optimal solutions on these approximations as the sample size increases or more data becomes available. Due to the weak convergence properties of sequences of probability measures, there is no guarantee that these quantities will exhibit favorable asymptotic properties. We consider a class of infinite-dimensional stochastic optimization problems inspired by recent work on PDE-constrained optimization as well as functional data analysis. For this class of problems, we provide both qualitative and quantitative stability results on the optimal value and optimal solutions. In both cases, we make use of the method of probability metrics. The optimal values are shown to be Lipschitz continuous with respect to a minimal information metric and consequently, under further regularity assumptions, with respect to certain Fortet-Mourier and Wasserstein metrics. We prove that even in the most favorable setting, the solutions are at best Hölder continuous with respect to changes in the underlying measure. The theoretical results are tested in the context of Monte Carlo approximation for a numerical example involving PDE-constrained optimization under uncertainty.

Convergence Analysis on Data-Driven Fortet-Mourier Metrics with Applications in Stochastic Optimization

Convergence Analysis for Distributionally Robust Optimization and Equilibrium Problems

On quantitative stability in infinite-dimensional optimization under uncertainty

Convergence Analysis for Mathematical Programs with Distributionally Robust Chance Constraint

Complete F-Moment Convergence for Weighted Sums of WOD Arrays with Statistical Applications

Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations

On Inhomogeneous Infinite Products of Stochastic Matrices and Applications

Complete F-Moment Convergence for Maximal Randomly Weighted Sums of Arrays of Rowwise Widely Orthant Dependent Random Variables and Its Statistical Applications

Convergence analysis of stochastic higher-order majorization-minimization algorithms

On Inhomogeneous Infinite Products of Stochastic Matrices and Their Applications

Complete F -Moment Convergence for a Class of Random Variables with Related Statistical Applications

Convergence Analysis for General Probability Flow ODEs of Diffusion Models in Wasserstein Distances

A General Result on Complete F-Moment Convergence with Its Application to Nonparametric Regression Models

Complete F-Moment Convergence for Widely Orthant Dependent Random Variables and Its Application in Nonparametric Models

Wasserstein Convergence Rate for Empirical Measures of Markov Processes

Complete $F$-Moment Convergence for Randomly Weighted Sums of Extended Negatively Dependent Random Variables and Its Statistical Application

Complete moment convergence for weighted sums of widely orthant-dependent random variables and its application in nonparametric regression models

Complete Convergence for Randomly Weighted Sums of Random Variables Satisfying Some Moment Inequalities

Global Optimization with Orthogonality Constraints Via Stochastic Diffusion on Manifold

On quantitative convergence for stochastic processes: Crossings, fluctuations and martingales

Convergence analysis of some multivariate Markov chains using stochastic monotonicity