Abstract:This paper proposes a second-order conic programming (SOCP) approach to solvedistributionally robust two-stage stochastic linear programs over 1-Wassersteinballs. We start from the case with distribution uncertainty only in theobjective function and exactly reformulate it as an SOCP problem. Then, westudy the case with distribution uncertainty only in constraints, and show thatsuch a robust program is generally NP-hard as it involves a norm maximizationproblem over a polyhedron. However, it is reduced to an SOCP problem if theextreme points of the polyhedron are given as a prior. This motivates to designa constraint generation algorithm with provable convergence to approximatelysolve the NP-hard problem. In sharp contrast to the exiting literature, thedistribution achieving the worst-case cost is given as an "empirical"distribution by simply perturbing each sample for both cases. Finally,experiments illustrate the advantages of the proposed model in terms of theout-of-sample performance and the computational complexity.

Second-Order Conic Programming Approach for Wasserstein Distributionally Robust Two-Stage Linear Programs