Quantifying and minimizing the risk is a basic problem faced in a wide range of applications. Once the risk is explicitly quantified by a risk measure, the crucial and ambitious goal is to obtain risk-averse solutions, given the computational hurdle typically associated with opti- mization problems under risk. This is especially true for many difficult combinatorial problems, and notably for scheduling problems. This paper aims to present a few tractable risk measures for the selective scheduling problem with parallel identical machines and sequence-dependent setup times. We indicate how deterministic reformulations can be obtained when the distributional information is limited to first and second-order moment information for a broad class of risk measures. We propose an efficient heuristic for addressing the computational difficulty of the resulting models and we showcase the practical applicability of the pro- posed approach providing computational evidence on a set of benchmark instances.
Tractable Risk Measures for the Selective Scheduling Problem with Sequence-Dependent Setup Times / Bruni, M. E.; Khodaparasti, S. - ELETTRONICO. - 1162:(2020), pp. 70-84. (Intervento presentato al convegno International Conference on Operations Research and Enterprise Systems) [10.1007/978-3-030-37584-3_4].
Tractable Risk Measures for the Selective Scheduling Problem with Sequence-Dependent Setup Times
Bruni M. E.;Khodaparasti S
2020
Abstract
Quantifying and minimizing the risk is a basic problem faced in a wide range of applications. Once the risk is explicitly quantified by a risk measure, the crucial and ambitious goal is to obtain risk-averse solutions, given the computational hurdle typically associated with opti- mization problems under risk. This is especially true for many difficult combinatorial problems, and notably for scheduling problems. This paper aims to present a few tractable risk measures for the selective scheduling problem with parallel identical machines and sequence-dependent setup times. We indicate how deterministic reformulations can be obtained when the distributional information is limited to first and second-order moment information for a broad class of risk measures. We propose an efficient heuristic for addressing the computational difficulty of the resulting models and we showcase the practical applicability of the pro- posed approach providing computational evidence on a set of benchmark instances.File | Dimensione | Formato | |
---|---|---|---|
Tractable.pdf
accesso riservato
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Non Pubblico - Accesso privato/ristretto
Dimensione
254.74 kB
Formato
Adobe PDF
|
254.74 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2980521