1: \begin{abstract}In this study, we consider two classes of multicriteria

2: two-stage stochastic programs in finite probability spaces with

3: multivariate risk constraints. The first-stage problem features a

4: multivariate stochastic benchmarking constraint based on a

5: vector-valued random variable representing multiple and possibly

6: conflicting stochastic performance measures associated with the

7: second-stage decisions. In particular, the aim is to ensure that the

8: associated random outcome vector of interest is preferable to a

9: specified benchmark with respect to the multivariate polyhedral

10: conditional value-at-risk (CVaR) or a multivariate stochastic order

11: relation. In this case, the classical decomposition methods cannot

12: be used directly due to the complicating multivariate stochastic

13: benchmarking constraints. We propose an exact unified decomposition

14: framework for solving these two classes of optimization problems and

15: show its finite convergence. We apply the proposed approach to a

16: stochastic network design problem in a pre-disaster humanitarian

17: logistics context and conduct a computational study concerning the

18: threat of hurricanes in the Southeastern part of the United States.

19: Our numerical results on these large-scale problems show that our

20: proposed algorithm is computationally scalable.

21: 

22: \ignore{our proposed algorithm tam precise degil ama abstract boyle

23: kalsa da olur...iki algoritma var ve asil SD-based olan scalable}

24: \end{abstract}

25: