Annals of Operations Research最新文献

Decision-making in complex systems often relies on machine learning models, yet highly accurate models such as XGBoost and neural networks can obscure the reasoning behind their predictions. In operations research applications, understanding how a decision is made is often as crucial as the decision itself. Traditional interpretable models, such as decision trees and logistic regression, provide transparency but may struggle with datasets containing intricate feature interactions. However, complexity in decision-making stems from interactions that are only relevant within certain subsets of data. Within these subsets, feature interactions may be simplified, forming simple structures where simple interpretable models can perform effectively. We propose a bottom-up simple structure-identifying algorithm that partitions data into interpretable subgroups known as simple structures, where feature interactions are minimized, allowing simple models to be trained within each subgroup. We demonstrate the robustness of the algorithm on synthetic data and show that the decision boundaries derived from simple structures are more interpretable and aligned with the intuition of the domain than those learned from a global model. By improving both explainability and predictive accuracy, our approach provides a principled framework for decision support in applications where model transparency is essential.

Queuing in vehicle routing problems happens when a given node requires to be visited by several vehicles, whereas only a limited number of vehicles can perform the service simultaneously. Hence, some vehicles must wait until the node is available. We present in this paper a mathheuristic approach to solve the problem. This approach incorporates two phases. The first phase executes a rolling horizon heuristic multiple times to generate an initial set of solutions. Those generated solutions are used to initialize a pool of routes. In the second phase, a column-generation based procedure is used to generate new routes. The contribution of our paper can be summarized as follows. (1) We implemented an efficient set partitioning model that allocates pre-determined slots of time to service operations of vehicles. (2) We proposed fast pricing heuristics to generate new routes with negative reduced costs. (3) The newly generated routes are based on existing ones, keeping the same physical description but the starting times of service operations are modified to better fit the queuing aspects. Performance evaluation has been conducted using instances derived from data provided by forest companies. Experiments proved the effectiveness of the proposed approach, by recording low route duration and achieving almost zero queuing times compared to the initial pool of solutions.

This paper uses a novel data envelopment analysis approach to assess the performance of supply chains in agriculture. Our network model treats the overall production and distribution process as consisting of separate stages. We account for the environmental externalities in agriculture and consider the links between different stages of the multi-stage process approximating agricultural supply chains. Additionally, we propose a new approach for modeling the impact of intermediate goods and calculate their marginal products. The empirical illustration is based on a panel of 18 cities in China’s Henan province during the period 2008–2021. Our results provide evidence of a significant potential for reducing carbon dioxide emission and increasing the agricultural output reaching final consumers in Henan. Notably, we find that overproduction of intermediate products can have an inhibitory effect on the final output, underscoring the importance of optimal resource allocation within the agricultural supply chain. We believe our findings can provide valuable insights for a more sustainable management of agricultural resources.

This study examines the purchasing decision-making of retail pharmacies when the potential for emergencies arises within a single cycle. Beyond accounting for conventional demand, retail pharmacies also need to plan for emergency demand. This paper employs the classic newsvendor model as a benchmark (PN) and explores three pre-purchasing strategies: a combination of conventional procurement and option procurement (POM), one-time procurement taking possible emergencies into account (PNO), a combination of conventional procurement and emergency procurement (PNE). Through an analysis of these procurement strategies, we find that, while POM usually performs better as a strategy, its position is affected by emergency shortage cost, exercise price, inventory cost, and the timing of emergency situations. Especially, exercise price changes does not always benefit retail pharmacies. Furthermore, neither PNO nor PNE provide any absolute advantages. PNO performs excellently when faced with higher emergency wholesale prices or lower emergency shortage costs. On the contrary, under certain conditions, PNE becomes a favorable choice for retail pharmacies. Specifically, when the emergency demand of retail pharmacies increases significantly, PNE is most suitable for retail pharmacies, while PNO is more advantageous when emergency situations occur near the end of the cycle.

Sensibly planning the subjects to study during a university degree is one of the most crucial tasks that impact the future professional life of a student. Nonetheless, to the best of our knowledge, no automated solution is available for students who want to plan their desired degree path and maximize the skills required by desired or target job(s). In this paper, we consider the Degree Planning Problem (DPP), which aims at computing degree plans composed of university subjects for students during the completion of an undergraduate degree. Specifically, we aim to obtain the best set of skills matching the requirements of students’ preferred job(s). To achieve this objective, we propose a flexible and scalable approach that solves the DPP in real-time by means of a non-trivial formalization as an optimization problem that can be solved with standard solvers. Finally, we employ real data from our University’s Bachelor in Information and Communications Technology to show, through several use cases, that our approach can be a valuable decision-support tool for students and curriculum designers.

Evidence of the impact of climate change has increased in recent years, forcing supply chain management (SCM) entities to promote sustainability in their operations. Meanwhile, global nations promote their sustainable development goals (SDGs) as one of the most important indicators to achieve sustainability regardless of application. SCM is a complex phenomenon of decision-making, which becomes even more complex when SDGs are adopted in SCM. Despite the central importance of the topic, only a small number of studies have reported on decision-making methods to improve the adoption of SDGs in SCM. To explore the state-of-the-art knowledge of operations research methods (ORMs) and models, we present this important Annals of Operations Research (ANOR) special issue. To introduce the special issue, we review the existing literature on models and methods to promote sustainability in SCM effectively and establish some associated future research directions. The findings of this editorial paper, as well as the articles in the special issue, can be used by both researchers and practitioners to consolidate recent advances and practices for establishing ORMs and models in SCM for effective decision-making to adopt SDGs. Additionally, this special issue paves the way for a solid foundation of advanced methods.

Bunkering costs constitute the largest portion of operational expenses in the shipping industry, directly influencing both economic efficiency and environmental impact. In line with Sustainable Development Goals (SDGs) 12 and 13, this study develops a decision model that jointly optimizes bunkering and sailing speed in liner shipping, with the aim of minimizing fuel costs and reducing carbon emissions. The model explicitly incorporates two often-overlooked aspects of the bunkering process: the requirement for empty tanks before refueling and the fuel inspection process. Due to the presence of infinite-dimensional and nonlinear terms, solving the model is computationally challenging. To address this complexity, we employ approximation algorithms and linearization techniques to transform the model into a mixed-integer linear programming (MILP) formulation. Additionally, we implement a Branch-and-Cut algorithm to enhance computational efficiency. Numerical experiments are conducted to evaluate the model’s performance, along with sensitivity analyses to assess the impact of key parameters. The results demonstrate that both the empty tank requirement and fuel inspection significantly influence bunkering decisions and sailing strategies, with the latter having a more pronounced effect. Moreover, our findings highlight the potential for sustainable fuel management practices to contribute to carbon reduction in maritime transportation. This study provides valuable insights for policymakers and industry stakeholders seeking to balance cost efficiency and environmental sustainability in shipping operations.

This paper investigates the impacts of trading behaviors on price discovery in futures markets. A dynamical model with difference equations is proposed to depict the interactions of heterogenous investors and the spot-futures coevolution. The system equilibrium and its stability conditions are mathematically analyzed. In the equilibrium, the futures price and the spot price converge to the fundamental value simultaneously. Stability conditions are necessary for the convergence process as well as the price discovery function. To ensure stability conditions, factors such as investor bounded rationality, risk appetites and market liquidity need to satisfy specific relationships. As the findings show, the arbitrage is not always beneficial to market stability and price discovery. It may increase price fluctuations in some cases. If investors have high degree of rationality, they tend to switch trading strategies with high intensity, which may destabilize the market. The simulations suggest the occurrence of complicated dynamics when stability conditions are violated. It provides theoretical insights into complicated phenomena in futures markets.

This paper proposes a game-theoretic model to investigate store brand (SB) introduction and information sharing decisions within a co-opetitive supply chain between a national brand (NB) manufacturer and a retailer, where the retailer chooses whether to introduce a store brand and share demand information with the NB manufacturer. In the main model, we consider that the retailer’s store brand introduction decision precedes the information sharing decision. We fully explore four scenarios involving the retailer’s strategic decision characteristics: no store brand introduction and no information sharing (NN), no store brand introduction and information sharing (NS), store brand introduction and no information sharing (IN), and store brand introduction and information sharing (IS). Our analysis draws the following interesting conclusions: Firstly, our research reveals that the NB manufacturer will benefit from SB introduction under certain conditions. Secondly, we demonstrate that the retailer’s store brand introduction and information sharing reinforce each other. Finally, our study also makes a contribution to how the degree of SB spillover and the fixed introduction cost affect the equilibrium strategy outcome.