Software

Overview
Publications
Team

Stochastic Decomposition
Stochastic decomposition (SD) is a sequential sampling-based algorithm for two-stage stochastic linear programs (2-SLP). The algorithmic details can be found in:

  1. Higle, J. L. and Sen, S. (1994). Finite master programs in regularized stochastic decomposition. Mathematical Programming, 67(1):143–168.
  2. Sen, S. and Liu, Y. (2016). Mitigating uncertainty via compromise decisions in two-stage stochastic linear programming: Variance reduction. Operations Research, 64(6):1422–1437.

The original version of the software was developed by Jason Mai, Lei Zhao, Yifan Liu, Harsha Gangammanavar, and Suvrajeet Sen. It is available at the Github repository of USC 3D Laboratory (https://github.com/USC3DLAB/SD) . Harsha Gangammanavar maintains the current version.

v1.0 : Supports 2-SLPs with randomness in right-hand side and the technology matrix.
v2.0 : Supports 2-SLPs with randomness in right-hand side vector, the technology matrix and cost coefficient vector.

Available on GitHub ( https://github.com/SMU-SODA/stochasticDecomposition )
Distributionally Robust Stochastic Decomposition
Distributionally Robust Stochastic decomposition (DRSD) is a sequential sampling-based algorithm for two-stage distributionally robust linear optimization. The algorithmic details can be found in:

Harsha Gangammanavar develops this software.


Available on GitHub (https://github.com/SMU-SODA/distributionallyRobust)
Mailing Address

P.O. Box 750123, Dallas, TX 75275-0123

Email Address

harsha {at} smu {dot} edu

Phone

214-768-3076