Show simple item record

dc.contributor.authorDe Moor, Bram J.
dc.contributor.authorGijsbrechts, Joren
dc.contributor.authorBoute, Robert
dc.date.accessioned2021-11-03T09:31:55Z
dc.date.available2021-11-03T09:31:55Z
dc.date.issued2021en_US
dc.identifier.issn0377-2217
dc.identifier.doi10.1016/j.ejor.2021.10.045
dc.identifier.urihttp://hdl.handle.net/20.500.12127/6986
dc.description.abstractDeep reinforcement learning (DRL) has proven to be an effective, general-purpose technology to develop ‘good’ replenishment policies in inventory management. We show how transfer learning from existing, well-performing heuristics may stabilize the training process and improve the performance of DRL in inventory control. While the idea is general, we specifically implement potential-based reward shaping to a deep Q-network algorithm to manage inventory of perishable goods that, cursed by dimensionality, has proven to be notoriously complex. The application of our approach may not only improve inventory cost performance and reduce computational effort, the increased training stability may also help to gain trust in the policies obtained by black box DRL algorithms.en_US
dc.language.isoenen_US
dc.publisherElsevieren_US
dc.subjectInventoryen_US
dc.subjectPerishable Inventory Managementen_US
dc.subjectDeep Reinforcement Learningen_US
dc.subjectReward Shapingen_US
dc.subjectTransfer Learningen_US
dc.titleReward shaping to improve the performance of deep reinforcement learning in perishable inventory managementen_US
dc.identifier.journalEuropean Journal of Operational Researchen_US
dc.contributor.departmentResearch Center for Operations Management, KU Leuven, Naamsestraat 69, Box 3555, 3000 Leuven, Belgiumen_US
dc.contributor.departmentIESEG School of Management, Rue de la Digue 3, 59000 Lille, Franceen_US
dc.contributor.departmentCatólica Lisbon School of Business and Economics, Palma de Cima, 1649-023 Lisbon, Portugalen_US
vlerick.knowledgedomainOperations & Supply Chain Managementen_US
vlerick.typearticleVlerick strategic journal articleen_US
vlerick.vlerickdepartmentTOMen_US
dc.identifier.vperid102358en_US


Files in this item

Thumbnail
Name:
Publisher version

This item appears in the following Collection(s)

Show simple item record