本文精选了运筹学国际顶刊《Operations Research》近期发表的论文,提供运筹学领域最新的学术动态。
Preservation of Multimodularity: New Results and Applications
原刊和作者:
Operations Research, Volume 74, Issue 5
Tong Wang (City University of Hong Kong)
Li Xiao (University of Macau)
Fen Xu (Huazhong University of Science and Technology)
Abstract
This paper establishes two new preservation results of multimodularity for two classes of resource allocation problems, where multiple resources can be used to satisfy multiple demands. Our results show that if allocation priorities between multiple resources and demands are determined by several marginal cost/value inequalities of the objective function, then the multimodularity and marginal cost/value inequalities are both preserved after optimization. We demonstrate their applications to several stochastic optimization problems, including dynamic capacity management problem with general upgrading, remanufacturing inventory systems with multiple return types under the lost-sales setting, buyer-item matching on resale platforms, and dynamic transshipment across multiple locations.
Link: https://doi.org/10.1287/opre.2025.1704
An Online Mirror Descent Learning Algorithm for Multiproduct Inventory Systems
原刊和作者:
Operations Research, Volume 74, Issue 5
Sichen Guo (University of Miami)
Cong Shi (University of Miami)
Chaolin Yang (Shanghai University of Finance and Economics)
Christos Zacharias (University of Miami)
Abstract
We study a canonical inventory control problem: a multiproduct, periodic-review, lost-sales inventory system with a warehouse-capacity constraint. We study this well-researched problem under the lens of demand learning from censored data. Unlike the traditional literature, we do not assume that demand distributions are known a priori. Instead, the decision maker only has access to observed sales data, whereas the lost-sales quantity remains unobserved. Existing online learning algorithms bear limitations in providing good-quality solutions to inventory systems offering a large variety of products. We employ and innovate mirror descent with cyclic update techniques to address the challenge of high dimensionality in product menus. We prove theoretically that our algorithm’s regret bound exhibits a logarithmic dependence on the number of products. This constitutes a significant improvement compared with the square-root regret bound established in the existing literature. Using empirical data, we implemented our methods to assess their practical merit and expose additional managerial insights. Our numerical study confirms that our methodology indeed produces inventory policies superior to existing state-of-the-art solutions, especially when managing a large menu of products. Drawing from our numerical observations and theory-informed insights, we provide clear guidelines for practical implementation along with fine-tuning recommendations.
Link: https://doi.org/10.1287/opre.2024.0982
Last-Mile Delivery of Malaria Prevention Products in the DRC: An Inventory Management Model Under Supply Chain Disruptions
Operations Research, Volume 74, Issue 5
Baris Ata (The University of Chicago)
Amy Lehman (Lake Tanganyika Floating Health Clinic)
Robert John Montgomery (University of California)
Abstract
This paper studies inventory management policies to improve the last-mile delivery of healthcare products in remote settings prone to supply chain disruptions. It is motivated by the need for a reliable and cost-effective delivery system to resupply new mosquito-repellent products to combat malaria in the Lake Tanganyika region of the Democratic Republic of the Congo. The primary delivery methods used to supply villages along Lake Tanganyika can become inoperable during periods of flooding, machinery breakdowns, or armed conflict. To ensure continuous coverage, our model allows for secondary delivery methods with per unit costs and no fixed setup costs to be used during disruptions. In normal operating periods, our model allows for delivery using a large-capacity, fixed-setup-cost method with no variable costs. We model the changes in the operating environment between normal and disrupted states using a time-varying Markov chain, which allows for the incorporation of seasonal effects, namely, a distinct rainy season. Using an infinite-horizon, discounted-cost objective, we show that an (??,??) policy is optimal for managing inventory when in the normal operating state. In an extensive numerical study, we observe that a hybrid policy comprising an (??,??) policy in the normal state and base-stock policies in the disrupted states is optimal in the great majority of settings considered. We also derive several simpler heuristic policies and identify the parameter regimes in which they are effective substitutes for the hybrid and optimal policies.
Link: https://doi.org/10.1287/opre.2022.0177
Planning a Community Approach to Diabetes Care in Low- and Middle-Income Countries Using Optimization
原刊和作者:
Operations Research, Volume 74, Issue 5
Katherine B. Adams (University of Texas at San Antonio)
Justin J. Boutilier (University of Ottawa)
Sarang Deo (Indian School of Business)
Yonatan Mintz (University of Wisconsin–Madison)
Abstract
Diabetes is a global health priority, especially in low- and middle-income countries, where over 50% of premature deaths are attributed to high blood glucose. Community health worker (CHW) programs can provide affordable and culturally tailored solutions for early detection and management of diabetes. We introduce an optimization framework to determine personalized CHW visits that maximize glycemic control at a community level. Our framework explicitly models the trade-off between screening new patients and providing management visits to individuals who are enrolled in treatment. We account for patients’ motivational states, which affect their decisions to enroll or drop out of treatment and therefore, the effectiveness of the intervention. By estimating patients’ health and motivational states, our model builds visit plans accounting for patients’ trade-offs when deciding to enroll in treatment, leading to reduced dropout rates and improved resource allocation. We apply our approach to generate CHW visit plans using operational data from urban slums in India. We find that our approach can reduce fasting blood glucose by up to 25% with the same capacity as the best baseline method. Our experiments also demonstrate that our approach performs well with imperfect information.
Link: https://doi.org/10.1287/opre.2023.0248
Search Games with Predictions
Operations Research, Volume 74, Issue 5
Spyros Angelopoulos (CNRS and Sorbonne Université)
Thomas Lidbetter (Rutgers University)
Konstantinos Panagiotou (LMU Munich)
Abstract
We introduce the study of search games between a mobile Searcher and an immobile Hider in a new setting in which the Searcher has some potentially erroneous information, or prediction, on the Hider’s position. The objective is to establish tight tradeoffs between the consistency of a search strategy (i.e., its worst-case expected payoff assuming the prediction is correct) and its robustness (i.e., the worst-case expected payoff without any assumptions on the quality of the prediction). Our study is the first in the realm of learning-augmented algorithms to address the full power of mixed (randomized) strategies; previous work focused only on deterministic strategies, or relied on stochastic assumptions that do not guarantee worst-case robustness in adversarial situations. We provide a framework for proving the optimality of consistency/robustness tradeoffs in search games over both discrete and continuous as well as bounded and unbounded spaces. We illustrate the approach using three well-known problems, namely, searching in discrete locations, searching with stochastic overlook, and searching in the infinite line. Our techniques are not limited to search games, but they can be applied, more broadly, to any two-person zero-sum games in learning-augmented settings.
Link: https://doi.org/10.1287/opre.2024.1498