On parallelizing dual decomposition in stochastic integer programming

For stochastic mixed-integer programs, we revisit the dual decomposition algorithm of Car\o{}e and Schultz from a computational perspective with the aim of its parallelization. We address an important bottleneck of parallel execution by identifying a formulation that permits the parallel solution of the \textit{master} program by using structure-exploiting interior-point solvers. Our results demonstrate the potential for parallel speedup and the importance of regularization (stabilization) in the dual optimization. Load imbalance is identified as a remaining barrier to parallel scalability.


Accepted for publication in Operations Research Letters. Formerly preprint ANL/MCS-P3037-0912, Argonne National Laboratory, October 2012.