Risk-averse stochastic modeling and optimization

Noyan, Nilay (2018) Risk-averse stochastic modeling and optimization. In: Gel, Esma and NTAIMO, LEWIS, (eds.) Recent Advances in Optimization and Modeling of Contemporary Problems. INFORMS (Institute for Operations Research and Management Sciences), Catonsville, MD, USA, pp. 221-254. ISBN 978-0-9906153-2-3

[img]PDF (INFORMS Tutorial Chapter) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader

Official URL: http://dx.doi.org/10.1287/educ.2018.0183


The ability to compare random outcomes based on the decision makers' risk preferences is crucial to modeling decision-making problems under uncertainty. In this tutorial, the primary focus is on the stochastic preference relations based on the widely applied risk measure, conditional value-at-risk (CVaR), and the second-order stochastic dominance (SSD). We present single- and two-stage stochastic optimization problems that feature such risk-averse preference relations. We discuss the main computational challenges in solving the problems of interest, and for finite probability spaces, we describe alternative mathematical programming formulations and effective solution methods. Our focus is on delayed cut generation solution algorithms, which rely on a Benders-type scenario decomposition approach in the case of a two-stage problem. In addition, we review the recent developments in risk-averse stochastic programming, with a particular focus on multicriteria optimization problems that feature multivariate stochastic benchmarking constraints based on CVaR and SSD.

Item Type:Book Section / Chapter
Additional Information:
Subjects:Q Science > Q Science (General)
ID Code:37495
Deposited By:Nilay Noyan
Deposited On:07 Aug 2019 15:24
Last Modified:07 Aug 2019 15:24

Repository Staff Only: item control page