Adaptive and Nonadaptive Samples in Solving Stochastic Linear Programs: A Computational Investigation

Large scale stochastic linear programs are typically solved using a combination of mathematical programming techniques and sample-based approximations. Some methods are designed to permit sample sizes to adapt to information obtained during the solution process, while others are not. In this paper, we experimentally examine the relative merits of approximations based on adaptive samples and those based on non-adaptive samples. We begin with an examination of two versions of an adaptive technique, Stochastic Decomposition (SD), and conclude with a comparison to a nonadaptive technique, the Sample Average Approximation method (SAA). Our results indicate that there is minimal difference in the quality of the solutions provided by SD and SAA, although SAA requires substantially more time to execute. Acknowledgement: This work was supported by Grant No. DMS 04-00085 from The National Science Foundation. [email protected]