- paper_1_arxiv
Accepted author manuscript, 572 KB, PDF document

Available under license: CC BY: Creative Commons Attribution 4.0 International License

- http://link.springer.com/article/10.1007/s10107-019-01451-7
Final published version

Research output: Contribution to journal › Journal article

E-pub ahead of print

**Problem-driven scenario generation : an analytical approach for stochastic programs with tail risk measure.** / Fairbrother, Jamie; Turner, Amanda; Wallace, Stein W.

Research output: Contribution to journal › Journal article

Fairbrother, J, Turner, A & Wallace, SW 2019, 'Problem-driven scenario generation: an analytical approach for stochastic programs with tail risk measure', *Mathematical Programming*. https://doi.org/10.1007/s10107-019-01451-7

Fairbrother, J., Turner, A., & Wallace, S. W. (2019). Problem-driven scenario generation: an analytical approach for stochastic programs with tail risk measure. *Mathematical Programming*. https://doi.org/10.1007/s10107-019-01451-7

Fairbrother J, Turner A, Wallace SW. Problem-driven scenario generation: an analytical approach for stochastic programs with tail risk measure. Mathematical Programming. 2019 Nov 26. https://doi.org/10.1007/s10107-019-01451-7

@article{f62cf21fe6004f60bd8acb52da08b982,

title = "Problem-driven scenario generation: an analytical approach for stochastic programs with tail risk measure",

abstract = "Scenario generation is the construction of a discrete random vector to represent parameters of uncertain values in a stochastic program. Most approaches to scenario generation are distribution-driven, that is, they attempt to construct a random vector which captures well in a probabilistic sensethe uncertainty. On the other hand, a problem-driven approach may be able to exploit the structure of a problem to provide a more concise representation of the uncertainty.In this paper we propose an analytic approach to problem-driven scenario generation. This approach applies to stochastic programs where a tail risk measure, such as conditional value-at-risk, is applied to a loss function. Since tail risk measures only depend on the upper tail of a distribution, standard methods of scenario generation, which typically spread their scenarios evenly across the support of the random vector, struggle to adequately represent tail risk. Our scenario generation approach works by targeting the construction of scenarios in areas of the distribution corresponding to the tails of the loss distributions. We provide conditions under which our approach is consistentwith sampling, and as proof-of-concept demonstrate how our approach could be applied to two classes of problem, namely network design and portfolio selection. Numerical tests on the portfolio selection problem demonstratethat our approach yields better and more stable solutions compared to standard Monte Carlo sampling.",

author = "Jamie Fairbrother and Amanda Turner and Wallace, {Stein W.}",

year = "2019",

month = "11",

day = "26",

doi = "10.1007/s10107-019-01451-7",

language = "English",

journal = "Mathematical Programming",

issn = "0025-5610",

publisher = "Springer-Verlag GmbH and Co. KG",

}

TY - JOUR

T1 - Problem-driven scenario generation

T2 - an analytical approach for stochastic programs with tail risk measure

AU - Fairbrother, Jamie

AU - Turner, Amanda

AU - Wallace, Stein W.

PY - 2019/11/26

Y1 - 2019/11/26

N2 - Scenario generation is the construction of a discrete random vector to represent parameters of uncertain values in a stochastic program. Most approaches to scenario generation are distribution-driven, that is, they attempt to construct a random vector which captures well in a probabilistic sensethe uncertainty. On the other hand, a problem-driven approach may be able to exploit the structure of a problem to provide a more concise representation of the uncertainty.In this paper we propose an analytic approach to problem-driven scenario generation. This approach applies to stochastic programs where a tail risk measure, such as conditional value-at-risk, is applied to a loss function. Since tail risk measures only depend on the upper tail of a distribution, standard methods of scenario generation, which typically spread their scenarios evenly across the support of the random vector, struggle to adequately represent tail risk. Our scenario generation approach works by targeting the construction of scenarios in areas of the distribution corresponding to the tails of the loss distributions. We provide conditions under which our approach is consistentwith sampling, and as proof-of-concept demonstrate how our approach could be applied to two classes of problem, namely network design and portfolio selection. Numerical tests on the portfolio selection problem demonstratethat our approach yields better and more stable solutions compared to standard Monte Carlo sampling.

AB - Scenario generation is the construction of a discrete random vector to represent parameters of uncertain values in a stochastic program. Most approaches to scenario generation are distribution-driven, that is, they attempt to construct a random vector which captures well in a probabilistic sensethe uncertainty. On the other hand, a problem-driven approach may be able to exploit the structure of a problem to provide a more concise representation of the uncertainty.In this paper we propose an analytic approach to problem-driven scenario generation. This approach applies to stochastic programs where a tail risk measure, such as conditional value-at-risk, is applied to a loss function. Since tail risk measures only depend on the upper tail of a distribution, standard methods of scenario generation, which typically spread their scenarios evenly across the support of the random vector, struggle to adequately represent tail risk. Our scenario generation approach works by targeting the construction of scenarios in areas of the distribution corresponding to the tails of the loss distributions. We provide conditions under which our approach is consistentwith sampling, and as proof-of-concept demonstrate how our approach could be applied to two classes of problem, namely network design and portfolio selection. Numerical tests on the portfolio selection problem demonstratethat our approach yields better and more stable solutions compared to standard Monte Carlo sampling.

U2 - 10.1007/s10107-019-01451-7

DO - 10.1007/s10107-019-01451-7

M3 - Journal article

JO - Mathematical Programming

JF - Mathematical Programming

SN - 0025-5610

ER -