Recovering best statistical guarantees via the empirical divergence-based distributionally robust optimization
From MaRDI portal
Publication:5129181
DOI10.1287/OPRE.2018.1786zbMATH Open1455.90122arXiv1605.09349OpenAlexW2962771675WikidataQ89699888 ScholiaQ89699888MaRDI QIDQ5129181FDOQ5129181
Authors: Henry Lam 
Publication date: 26 October 2020
Published in: Operations Research (Search for Journal in Brave)
Abstract: We investigate the use of distributionally robust optimization (DRO) as a tractable tool to recover the asymptotic statistical guarantees provided by the Central Limit Theorem, for maintaining the feasibility of an expected value constraint under ambiguous probability distributions. We show that using empirically defined Burg-entropy divergence balls to construct the DRO can attain such guarantees. These balls, however, are not reasoned from the standard data-driven DRO framework since by themselves they can have low or even zero probability of covering the true distribution. Rather, their superior statistical performances are endowed by linking the resulting DRO with empirical likelihood and empirical processes. We show that the sizes of these balls can be optimally calibrated using chi-square process excursion. We conduct numerical experiments to support our theoretical findings.
Full work available at URL: https://arxiv.org/abs/1605.09349
Recommendations
- Statistics of robust optimization: a generalized empirical likelihood approach
- scientific article; zbMATH DE number 7668294
- Distributionally robust optimization. A review on theory and applications
- Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations
- Frameworks and results in distributionally robust optimization
central limit theoremempirical processempirical likelihooddistributionally robust optimization\(\chi^2\)-process
Cites Work
- Weak convergence and empirical processes. With applications to statistics
- Fast simulation of Gaussian random fields
- Title not available (Why is that?)
- On Information and Sufficiency
- Bounds on coverage probabilities of the empirical likelihood ratio confidence regions.
- Empirical likelihood is Bartlett-correctable
- Empirical likelihood
- Empirical likelihood ratio confidence intervals for a single functional
- Empirical likelihood on the full parameter space
- A new method of calibration for the empirical loglikelihood ratio
- Recent advances in robust optimization: an overview
- Random Fields and Geometry
- Distributionally robust optimization under moment uncertainty with application to data-driven problems
- A Sample Approximation Approach for Optimization with Probabilistic Constraints
- Lectures on Stochastic Programming
- Efficient Monte Carlo for high excursions of Gaussian random fields
- Topological complexity of smooth random functions. École d'Été de Probabilités de Saint-Flour XXXIX-2009.
- Sample average approximation of expected value constrained stochastic programs
- Distributionally robust optimization and its tractable approximations
- Title not available (Why is that?)
- Data-driven chance constrained stochastic program
- Call center staffing with simulation and cutting plane methods
- Handling CVaR objectives and constraints in two-stage stochastic models
- Likelihood robust optimization for data-driven problems
- Adjusted empirical likelihood with high-order precision
- Finite-sample properties of the adjusted empirical likelihood
- Distributionally Robust Convex Optimization
- Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations
- The empirical likelihood approach to quantifying uncertainty in sample average approximation
- Robust empirical optimization is almost the same as mean-variance optimization
- Robust sensitivity analysis for stochastic systems
- Quantifying distributional model risk via optimal transport
- Robust sample average approximation
Cited In (21)
- Learning models with uniform performance via distributionally robust optimization
- Data-driven robust resource allocation with monotonic cost functions
- Distributionally robust bottleneck combinatorial problems: uncertainty quantification and robust decision making
- Distributionally robust resource planning under binomial demand intakes
- Statistics of robust optimization: a generalized empirical likelihood approach
- Risk and complexity in scenario optimization
- Distributionally Robust Stochastic Dual Dynamic Programming
- Optimization-based calibration of simulation input models
- Gradient and Hessian of joint probability function with applications on chance-constrained programs
- Robust Actuarial Risk Analysis
- Residuals-based distributionally robust optimization with covariate information
- Robust Simulation with Likelihood-Ratio Constrained Input Uncertainty
- Rejoinder: New Objectives for Policy Learning
- Worst-case moments under partial ambiguity
- Parametric scenario optimization under limited data: a distributionally robust optimization view
- Robust mean variance optimization problem under Rényi divergence information
- Distributionally robust optimization for sequential decision-making
- Data-driven remanufacturing planning with parameter uncertainty
- Robust Wasserstein profile inference and applications to machine learning
- Frameworks and results in distributionally robust optimization
- Title not available (Why is that?)
Uses Software
This page was built for publication: Recovering best statistical guarantees via the empirical divergence-based distributionally robust optimization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5129181)
