1: \begin{abstract}

2:  The two-stage program with the distribution uncertainty is a fundamental optimization problem with broad applications. This paper proposes a novel data-driven model over 1-Wasserstein balls to handle the distribution uncertainty. Moreover, it applies to the situation where the true distribution is slowly time-varying. An SOCP-based algorithm with provable convergence has been presented to solve the two-stage program over 1-Wasserstein balls. The proposed model is particularly effective for applications with uncertainty only in the objective functions or constraints. This paper focuses on the applications in two-stage portfolio programs and two-stage material order programs. The good out-of-sample performance and computational complexity of the model are validated by the experiments on these applications. This model can be easily applied to other applications, such as the two-stage schedule problems and recourse allocation problems.

3: 

4: \end{abstract}

5: