Sinkhorn Distributionally Robust Optimization
arXiv:2109.11926
Abstract
We study distributionally robust optimization with Sinkhorn distance -- a variant of Wasserstein distance based on entropic regularization. We derive a convex programming dual reformulation for general nominal distributions, transport costs, and loss functions. To solve the dual reformulation, we develop a stochastic mirror descent algorithm with biased subgradient estimators and derive its computational complexity guarantees. Finally, we provide numerical examples using synthetic and real data to demonstrate its superior performance.
55 pages, 15 figures
References in corpus (7)
- Distributionally Robust Stochastic Optimization with Dependence Structure
- A Distributionally Robust Approach to Fair Classification
- Distributionally Robust Reinforcement Learning
- Small-Sample Inferred Adaptive Recoding for Batched Network Coding
- A First-Order Algorithmic Framework for Wasserstein Distributionally Robust Logistic Regression
- Robustifying Conditional Portfolio Decisions via Optimal Transport
- Two-sample Test with Kernel Projected Wasserstein Distance