Decomposition Algorithm for Distributionally Robust Optimization using Wasserstein Metric

Abstract

We study distributionally robust optimization (DRO) problems where the ambiguity set is defined using the Wasserstein metric. We show that this class of DRO problems can be reformulated as semi-infinite programs. We give an exchange method to solve the reformulated problem for the general nonlinear model, and a central cutting-surface method for the convex case, assuming that we have a separation oracle. We used a distributionally robust generalization of the logistic regression model to test our algorithm. Numerical experiments on the distributionally robust logistic regression models show that the number of oracle calls are typically 20 ? 50 to achieve 5-digit precision. The solution found by the model is generally better in its ability to predict with a smaller standard error.

25 pages

Cited by in corpus (11)

• On Distributionally Robust Chance Constrained Programs with Wasserstein Distance
• Optimal Transport Based Distributionally Robust Optimization: Structural Properties and Iterative Schemes
• Tutorials on Advanced Optimization Methods
• Distributionally Robust Optimization with Decision Dependent Ambiguity Sets
• Practicable Robust Stochastic Optimization under Divergence Measures
• Tractable Reformulations of Distributionally Robust Two-stage Stochastic Programs with $\infty -$Wasserstein Distance
• Efficient Stochastic Gradient Descent for Learning with Distributionally Robust Optimization
• Wasserstein Distributionally Robust Inverse Multiobjective Optimization
• Robust Stochastic Optimization with Rare-Event Modeling
• Unbiased Gradient Estimation for Distributionally Robust Learning
• A Robust Learning Algorithm for Regression Models Using Distributionally Robust Optimization under the Wasserstein Metric