Efficient solution selection for two-stage stochastic programs

Xin Fei, Nalân Gülpınar and Jürgen Branke

European Journal of Operational Research, 2019, vol. 277, issue 3, 918-929

Abstract: Sampling-based stochastic programs are extensively applied in practice. However, the resulting models tend to be computationally challenging. A reasonable number of samples needs to be identified to represent the random data, and a group of approximate models can then be constructed using such a number of samples. These approximate models can produce a set of potential solutions for the original model. In this paper, we consider the problem of allocating a finite computational budget among numerous potential solutions of a two-stage linear stochastic program, which aims to identify the best solution among potential ones by conducting simulation under a given computational budget. We propose a two-stage heuristic approach to solve the computational resource allocation problem. First, we utilise a Wasserstein-based screening rule to remove potentially inferior solutions from the simulation. Next, we use a ranking and selection technique to efficiently collect performance information of the remaining solutions. The performance of our approach is demonstrated through well-known benchmark problems. Results show that our method provides good trade-offs between computational effort and solution performance.

Keywords: Stochastic programming; Sample average approximation; Wasserstein metric; Ranking and selection (search for similar items in EconPapers)
Date: 2019
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0377221719301432
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:ejores:v:277:y:2019:i:3:p:918-929

DOI: 10.1016/j.ejor.2019.02.015

Access Statistics for this article

European Journal of Operational Research is currently edited by Roman Slowinski, Jesus Artalejo, Jean-Charles. Billaut, Robert Dyson and Lorenzo Peccati

More articles in European Journal of Operational Research from Elsevier
Bibliographic data for series maintained by Catherine Liu ().