Enhancements of two-stage stochastic decomposition

Authors:
Suvrajeet Sen;Zhihong Zhou;Kai Huang
Affiliations:
The Data Driven Decisions Lab, Industrial and Systems Engineering, The Ohio State University, Columbus, OH 43210, USA and The MORE Institute, Department of Systems and Industrial Engineering, The ...;The MORE Institute, Department of Systems and Industrial Engineering, The University of Arizona, Tucson, AZ 85721, USA;School of Management, Binghamton University, The State University of New York, PO Box 6000, Binghamton, NY 13902-6000, USA
Venue:
Computers and Operations Research
Year:
2009

Citing 7
Cited 0

Scenarios and policy aggregation in optimization under uncertainty

Mathematics of Operations Research
Stochastic decomposition: an algorithm for two-state linear programs with recourse

Mathematics of Operations Research
On the convergence of algorithms with implications for stochastic and nondifferentiable optimization

Mathematics of Operations Research
Finite master programs in regularized stochastic decomposition

Mathematical Programming: Series A and B
Sample-path optimization of convex stochastic performance functions

Mathematical Programming: Series A and B
Decomposition Algorithms for Stochastic Programming on a Computational Grid

Computational Optimization and Applications
Monte Carlo bounding techniques for determining solution quality in stochastic programs

Operations Research Letters

Quantified Score

Hi-index	0.01

Visualization

Abstract

This paper presents some enhancements associated with stochastic decomposition (SD). Specifically, we study two issues: (a) Are there any conditions under which the regularized version of SD generates a unique solution? (b) Is there a way to modify the SD algorithm so that a user can trade-off solution times with solution quality? The second issue addresses the scalability of SD for very large scale problems for which computational resources may be limited and the user may be willing to accept solutions that are ''nearly optimal''. We show that by using bootstrapping (re-sampling) the regularized SD algorithm can be accelerated without significant loss of optimality. We report computational results that demonstrate the viability of this approach.