In this manuscript, we propose an innovative early warning Machine Learning-based model to identify potential threats to financial sustainability for non-financial companies. Unlike most state-of-the-art tools, whose outcomes are often difficult to understand even for experts, our model provides an easily interpretable visualization of balance sheets, projecting each company in a bi-dimensional space according to an autoencoder-based dimensionality reduction matched with a Nearest-Neighbor-based default density estimation. In the resulting space, the distress zones, where the default intensity is high, appear as homogeneous clusters directly identified. Our empirical experiments provide evidence of the interpretability, forecasting ability, and robustness of the bi-dimensional space.
{"title":"Corporate risk stratification through an interpretable autoencoder-based model","authors":"Alessandro Giuliani , Roberto Savona , Salvatore Carta , Gianmarco Addari , Alessandro Sebastian Podda","doi":"10.1016/j.cor.2024.106884","DOIUrl":"10.1016/j.cor.2024.106884","url":null,"abstract":"<div><div>In this manuscript, we propose an innovative early warning Machine Learning-based model to identify potential threats to financial sustainability for non-financial companies. Unlike most state-of-the-art tools, whose outcomes are often difficult to understand even for experts, our model provides an easily interpretable visualization of balance sheets, projecting each company in a bi-dimensional space according to an autoencoder-based dimensionality reduction matched with a Nearest-Neighbor-based default density estimation. In the resulting space, the distress zones, where the default intensity is high, appear as homogeneous clusters directly identified. Our empirical experiments provide evidence of the interpretability, forecasting ability, and robustness of the bi-dimensional space.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"174 ","pages":"Article 106884"},"PeriodicalIF":4.1,"publicationDate":"2024-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142593375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-30DOI: 10.1016/j.cor.2024.106879
Guiyu Li, Hongbo Duan
There are plenty of uncertainties in the integrated climate-economic system including parameter uncertainty and model uncertainty, which significantly challenges the assessment of climate goals committed in the Paris Agreement pledges. In this study, we develop a robustness assessment framework of climate policy by effectively coupling the distributionally robust optimization (DRO) methodology with integrated assessment models (IAMs), termed DRO-IAMS framework, where “S” emphasizes the multiple IAMs being incorporated. Our approach determines a safeguarding probability for the achievement of carbon-neutrality target through the worst-case Conditional Value-at-Risk (CVaR) criterion by effectively capturing the fat-tail effect and exploiting its tractability. Leveraging a discrete support of uncertain parameters over which the objective value of global temperature increase (GTI) can be readily accessible using the IAMs, our developed DRO-IAMS framework effectively circumvents the difficulty in utilizing analytically the black-box-featured IAMs, and achieves a comprehensive and more flexible fashion in integrating the DRO (e.g, moment, -divergence, and Wasserstein ambiguity sets) and IAMs (e.g., DICE, FUND, and E3METL) to cope with parameter- and model uncertainties in climate policy assessment. Our results suggest that parameter uncertainty and model uncertainty — as critical issues that can have significant impacts on the warming and economic performance of policies — could incur biased assessment for the realization of climate targets. Our proposed DRO-IAMS approach — by its design — is shown to be able to effectively mitigate such issues by pursuing stricter mitigation efforts, and can produce more reliable assessments for typical climate policies than the common sampling-based approaches.
{"title":"Robustness assessment of climate policies towards carbon neutrality: A DRO-IAMS approach","authors":"Guiyu Li, Hongbo Duan","doi":"10.1016/j.cor.2024.106879","DOIUrl":"10.1016/j.cor.2024.106879","url":null,"abstract":"<div><div>There are plenty of uncertainties in the integrated climate-economic system including parameter uncertainty and model uncertainty, which significantly challenges the assessment of climate goals committed in the Paris Agreement pledges. In this study, we develop a robustness assessment framework of climate policy by effectively coupling the distributionally robust optimization (DRO) methodology with integrated assessment models (IAMs), termed DRO-IAMS framework, where “S” emphasizes the multiple IAMs being incorporated. Our approach determines a safeguarding probability for the achievement of carbon-neutrality target through the worst-case Conditional Value-at-Risk (CVaR) criterion by effectively capturing the fat-tail effect and exploiting its tractability. Leveraging a discrete support of uncertain parameters over which the objective value of global temperature increase (GTI) can be readily accessible using the IAMs, our developed DRO-IAMS framework effectively circumvents the difficulty in utilizing analytically the black-box-featured IAMs, and achieves a comprehensive and more flexible fashion in integrating the DRO (<em>e.g</em>, moment, <span><math><mi>ϕ</mi></math></span>-divergence, and Wasserstein ambiguity sets) and IAMs (<em>e.g.</em>, DICE, FUND, and E3METL) to cope with parameter- and model uncertainties in climate policy assessment. Our results suggest that parameter uncertainty and model uncertainty — as critical issues that can have significant impacts on the warming and economic performance of policies — could incur biased assessment for the realization of climate targets. Our proposed DRO-IAMS approach — by its design — is shown to be able to effectively mitigate such issues by pursuing stricter mitigation efforts, and can produce more reliable assessments for typical climate policies than the common sampling-based approaches.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"174 ","pages":"Article 106879"},"PeriodicalIF":4.1,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142652146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Re-direction occurs when a customer arriving at a station in a queuing network has to be re-directed to a downstream station to complete service. Re-direction is extremely common in practice and occurs for a variety of reasons, ranging from incorrect initial station assignment to cases where the initial station only provides part of the service. Gatekeeper stations (e.g., information desks) is a special case of re-direction. We consider re-direction in a queueing network consisting of single-server stations serving two customer types with different service time requirements. The behavior of such queueing networks is quite complex: even when all external arrivals and all services are Markovian, the customers’ inter-departure distribution, and hence their arrival process to downstream stations, is non-Markovian. Thus, product-form representation does not hold for such networks. Our analysis focuses on the key building block: the inter-departure process from a station serving two distinct customer types and routing them to two different downstream service paths. Using a novel approach, we obtain a very accurate phase-type representation of the inter-departure process under equilibrium. We show that the resulting methodology has significant advantages over both simulation modeling (our method is much faster) and the available approximation techniques (our method is more accurate). Finally, we demonstrate an interesting phenomenon: even when the station merely re-directs one of the customer types (providing no service and seemingly useless waits), it can serve as a “regulator”, reducing the variability of the downstream arrival process. We show that, under some conditions, this can improve the overall system performance.
{"title":"Re-direction in queueing networks with two customer types: The inter-departure analysis","authors":"Opher Baron , Oded Berman , Dmitry Krass , Eliran Sherzer","doi":"10.1016/j.cor.2024.106867","DOIUrl":"10.1016/j.cor.2024.106867","url":null,"abstract":"<div><div>Re-direction occurs when a customer arriving at a station in a queuing network has to be re-directed to a downstream station to complete service. Re-direction is extremely common in practice and occurs for a variety of reasons, ranging from incorrect initial station assignment to cases where the initial station only provides part of the service. <em>Gatekeeper</em> stations (e.g., information desks) is a special case of re-direction. We consider re-direction in a queueing network consisting of single-server stations serving two customer types with different service time requirements. The behavior of such queueing networks is quite complex: even when all external arrivals and all services are Markovian, the customers’ inter-departure distribution, and hence their arrival process to downstream stations, is non-Markovian. Thus, product-form representation does not hold for such networks. Our analysis focuses on the key building block: the inter-departure process from a station serving two distinct customer types and routing them to two different downstream service paths. Using a novel approach, we obtain a very accurate phase-type representation of the inter-departure process under equilibrium. We show that the resulting methodology has significant advantages over both simulation modeling (our method is much faster) and the available approximation techniques (our method is more accurate). Finally, we demonstrate an interesting phenomenon: even when the station merely re-directs one of the customer types (providing no service and seemingly useless waits), it can serve as a “regulator”, reducing the variability of the downstream arrival process. We show that, under some conditions, this can improve the overall system performance.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"173 ","pages":"Article 106867"},"PeriodicalIF":4.1,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142572185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-28DOI: 10.1016/j.cor.2024.106881
Parisa Torabi , Ahmad Hemmati , Anna Oleynik , Guttorm Alendal
Covering Tour Problem (CTP) is a combinatorial optimization problem in which the objective is to identify a minimum-cost tour that satisfies the coverage of a certain subset of nodes in a graph. The Covering Tour Problem with Varying Coverage (CTP-VC) is an extension of this problem in which the coverage radius is dependent on the amount of time spent at each node. In this paper, we propose a novel approach to address the CTP-VC using a Deep Reinforcement Learning Hyperheuristic (DRLH). This study includes experiments on the existing Adaptive Metaheuristic to solve CTP-VC, to enhance its solution quality. Further, new heuristics and three selection methods, namely Uniform Random Selection (URS), adaptive Metaheuristic (AMH), and the proposed DRLH are introduced. We detail the computational setup, including the instance sets utilized, the training process for the DRLH agent, and the validation procedures for model selection. Through extensive experimentation and analysis, we evaluate the performance of different selection methods, assess the solution quality of the DRLH approach, investigate the robustness of selection methods, examine heuristic selection frequency, and analyze solution convergence. Our results demonstrate the efficacy of the DRLH approach in tackling the CTP-VC, offering promising insights for future research in the interface of combinatorial optimization and reinforcement learning methodologies.
{"title":"A deep reinforcement learning hyperheuristic for the covering tour problem with varying coverage","authors":"Parisa Torabi , Ahmad Hemmati , Anna Oleynik , Guttorm Alendal","doi":"10.1016/j.cor.2024.106881","DOIUrl":"10.1016/j.cor.2024.106881","url":null,"abstract":"<div><div>Covering Tour Problem (CTP) is a combinatorial optimization problem in which the objective is to identify a minimum-cost tour that satisfies the coverage of a certain subset of nodes in a graph. The Covering Tour Problem with Varying Coverage (CTP-VC) is an extension of this problem in which the coverage radius is dependent on the amount of time spent at each node. In this paper, we propose a novel approach to address the CTP-VC using a Deep Reinforcement Learning Hyperheuristic (DRLH). This study includes experiments on the existing Adaptive Metaheuristic to solve CTP-VC, to enhance its solution quality. Further, new heuristics and three selection methods, namely Uniform Random Selection (URS), adaptive Metaheuristic (AMH), and the proposed DRLH are introduced. We detail the computational setup, including the instance sets utilized, the training process for the DRLH agent, and the validation procedures for model selection. Through extensive experimentation and analysis, we evaluate the performance of different selection methods, assess the solution quality of the DRLH approach, investigate the robustness of selection methods, examine heuristic selection frequency, and analyze solution convergence. Our results demonstrate the efficacy of the DRLH approach in tackling the CTP-VC, offering promising insights for future research in the interface of combinatorial optimization and reinforcement learning methodologies.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"174 ","pages":"Article 106881"},"PeriodicalIF":4.1,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142577962","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-28DOI: 10.1016/j.cor.2024.106874
Pierre Hémono , Ahmed Nait Chabane , M’hammed Sahnoun
Collaborative robotics is becoming increasingly prevalent in industry 5.0, leading to a growing need to improve interactions and collaborations between humans and robots. However, the current approach to defining the sharing of responsibilities between humans and robots is empirical and uses the robot as an active fixture of parts, which is a sub-optimal method for establishing efficient collaboration. This article focuses on optimizing human–robot collaboration on an assembly line within the aerospace industry based on a real-world use case. The methodology adopted in this research entails employing the multi-objective optimization (MOO) method to effectively tackle both the reduction of makespan and the mitigation of working difficulty. Two techniques have been compared for implementation: the weighted sum and the -constraint methods, which allow the generation of solutions addressing multiple objectives simultaneously. The results offer chief robotics officers a new tool to design collaboration patterns between humans and robots, with practical implications for real industrial applications. This solution produces several results, including improving company competitiveness and productivity, while maintaining the central role of humans within the company and improving its well-being.
{"title":"Multi objective optimization of human–robot collaboration: A case study in aerospace assembly line","authors":"Pierre Hémono , Ahmed Nait Chabane , M’hammed Sahnoun","doi":"10.1016/j.cor.2024.106874","DOIUrl":"10.1016/j.cor.2024.106874","url":null,"abstract":"<div><div>Collaborative robotics is becoming increasingly prevalent in industry 5.0, leading to a growing need to improve interactions and collaborations between humans and robots. However, the current approach to defining the sharing of responsibilities between humans and robots is empirical and uses the robot as an active fixture of parts, which is a sub-optimal method for establishing efficient collaboration. This article focuses on optimizing human–robot collaboration on an assembly line within the aerospace industry based on a real-world use case. The methodology adopted in this research entails employing the multi-objective optimization (MOO) method to effectively tackle both the reduction of makespan and the mitigation of working difficulty. Two techniques have been compared for implementation: the weighted sum and the <span><math><mi>ɛ</mi></math></span>-constraint methods, which allow the generation of solutions addressing multiple objectives simultaneously. The results offer chief robotics officers a new tool to design collaboration patterns between humans and robots, with practical implications for real industrial applications. This solution produces several results, including improving company competitiveness and productivity, while maintaining the central role of humans within the company and improving its well-being.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"174 ","pages":"Article 106874"},"PeriodicalIF":4.1,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142577961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-28DOI: 10.1016/j.cor.2024.106885
Sheng-Long Jiang
Steelmaking and continuous casting scheduling problem (SCCSP) is a classic optimization problem increasingly incorporating more constraints, such as energy-related ones. However, classic evolutionary algorithms with “rigid” encoding schemes face challenges in finding optimal solutions for heavily constrained SCCSPs. Motivated by this gap, this paper first extends the mathematical model of the classic SCCSP to its variant under energy thresholds (ET-SCCSP) from both single- and multi-objective optimization perspectives, and derives several problem-specific properties. Next, this paper develops a solving algorithm named the soft encoding-based evolutionary algorithm (SoEA), which uses a real-valued vector to encode a feasible solution for SCCSPs. Furthermore, SoEA introduces the following components: (1) a peak-cutting backward list scheduling (PC-BLS) procedure to decode a real-valued vector into a feasible solution, and (2) a local search procedure to enhance the algorithm’s performance. Comparative results in the computational experiment demonstrate that the SoEA with the propose encoding/decoding scheme: (1) achieves better performance than exact solver for small-scale instances under energy thresholds, (2) obtains promising results for medium-scale instances compared to other schemes, and (3) can be intensified by the tailored local search procedure. The proposed SoEA can also serve as a benchmark or tutorial for the development and evaluation of high-efficiency algorithms for other SCCSPs with heavy constraints. The source code is available on the GitHub repository: https://github.com/janason/Soft-Scheduling/tree/master/SoEA.
{"title":"A soft encoding-based evolutionary algorithm for the steelmaking scheduling problem and its extension under energy thresholds","authors":"Sheng-Long Jiang","doi":"10.1016/j.cor.2024.106885","DOIUrl":"10.1016/j.cor.2024.106885","url":null,"abstract":"<div><div>Steelmaking and continuous casting scheduling problem (SCCSP) is a classic optimization problem increasingly incorporating more constraints, such as energy-related ones. However, classic evolutionary algorithms with “rigid” encoding schemes face challenges in finding optimal solutions for heavily constrained SCCSPs. Motivated by this gap, this paper first extends the mathematical model of the classic SCCSP to its variant under energy thresholds (ET-SCCSP) from both single- and multi-objective optimization perspectives, and derives several problem-specific properties. Next, this paper develops a solving algorithm named the soft encoding-based evolutionary algorithm (SoEA), which uses a real-valued vector to encode a feasible solution for SCCSPs. Furthermore, SoEA introduces the following components: (1) a peak-cutting backward list scheduling (PC-BLS) procedure to decode a real-valued vector into a feasible solution, and (2) a local search procedure to enhance the algorithm’s performance. Comparative results in the computational experiment demonstrate that the SoEA with the propose encoding/decoding scheme: (1) achieves better performance than exact solver for small-scale instances under energy thresholds, (2) obtains promising results for medium-scale instances compared to other schemes, and (3) can be intensified by the tailored local search procedure. The proposed SoEA can also serve as a benchmark or tutorial for the development and evaluation of high-efficiency algorithms for other SCCSPs with heavy constraints. The source code is available on the GitHub repository: <span><span>https://github.com/janason/Soft-Scheduling/tree/master/SoEA</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"174 ","pages":"Article 106885"},"PeriodicalIF":4.1,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142704701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study introduces an arc-flow formulation and the first branch-and-price-and-cut (BPC) algorithm designed to solve the bin-packing problem with fragile objects (BPPFO). This variant of the bin-packing problem originates in the field of telecommunications, particularly in the allocation of cellular calls to frequency channels. The arc-flow formulation is inspired by previous studies and modifies the graph construction method to accommodate fragility constraints. We proved the correctness of this formulation and demonstrated its superiority in instances with small maximum fragility through extensive experiments. The proposed BPC algorithm leverages advanced cutting and packing techniques and incorporates innovative elements such as problem reduction, additional cutting planes, and a label-setting-based exact pricing algorithm. The experimental results demonstrate that the proposed BPC algorithm is highly competitive with the state-of-the-art algorithm for solving the BPPFO and can successfully solve several previously unsolved instances.
{"title":"Arc-flow formulation and branch-and-price-and-cut algorithm for the bin-packing problem with fragile objects","authors":"Sunkanghong Wang, Shaowen Yao, Hao Zhang, Qiang Liu, Lijun Wei","doi":"10.1016/j.cor.2024.106878","DOIUrl":"10.1016/j.cor.2024.106878","url":null,"abstract":"<div><div>This study introduces an arc-flow formulation and the first branch-and-price-and-cut (BPC) algorithm designed to solve the bin-packing problem with fragile objects (BPPFO). This variant of the bin-packing problem originates in the field of telecommunications, particularly in the allocation of cellular calls to frequency channels. The arc-flow formulation is inspired by previous studies and modifies the graph construction method to accommodate fragility constraints. We proved the correctness of this formulation and demonstrated its superiority in instances with small maximum fragility through extensive experiments. The proposed BPC algorithm leverages advanced cutting and packing techniques and incorporates innovative elements such as problem reduction, additional cutting planes, and a label-setting-based exact pricing algorithm. The experimental results demonstrate that the proposed BPC algorithm is highly competitive with the state-of-the-art algorithm for solving the BPPFO and can successfully solve several previously unsolved instances.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"173 ","pages":"Article 106878"},"PeriodicalIF":4.1,"publicationDate":"2024-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553258","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-24DOI: 10.1016/j.cor.2024.106877
Farhana Huq , Nahar Sultana , Palash Roy , Md. Abdur Razzaque , Shamsul Huda , Mohammad Mehedi Hassan
Online food delivery (OFD) represents a rapidly evolving e-business application that leverages cloud computing data centers, playing a crucial role in meeting the demands of urban lifestyles. With diverse order fulfillment features and increasing expectations for service quality, the task of effectively assigning riders for timely long-distance, cross-regional deliveries presents a significant engineering challenge. Previous studies often relied on traditional rider allocation methods that fail to account for varying capacities, or they utilized non-intelligent systems that did not adequately address fluctuating order demands and service delays. In this study, we introduce a robust Mixed Integer Linear Programming (MILP) optimization framework designed to minimize the total service time and delivery cost for cross-regional orders. This framework divides a large OFD area into multiple regions and utilizes both transfer vehicles and riders to optimize deliveries. To enhance the predictive accuracy of our model, we incorporate advanced machine learning techniques. Specifically, we employ the Long Short-Term Memory (LSTM) model to forecast regional order demands accurately, reflecting the dynamic nature of the marketplace. Additionally, Extreme Gradient Boosting (XGBoost) is tailored to dynamically predict travel times from restaurants to customer locations, facilitating more precise scheduling and resource allocation within the MILP framework. These machine learning techniques significantly bolster the MILP framework by providing detailed, accurate predictions that improve decision-making processes and adaptability to real-time conditions. Acknowledging the complexity of this optimization problem, we further enhance our approach by integrating a meta-heuristic algorithm, Adaptive Large Neighbor Search (ALNS), which efficiently assigns orders to the appropriate transfer vehicles and riders within polynomial time. Our Cross Regional Online Food Delivery (XROFD) system is meticulously designed to optimize both customer satisfaction and rider incentives. Simulation experiments confirm that the XROFD system not only reduces service times and delivery costs but also markedly enhances customer satisfaction and provides superior incentives for riders, outperforming existing state-of-the-art methods.
{"title":"Cross regional online food delivery: Service quality optimization and real-time order assignment","authors":"Farhana Huq , Nahar Sultana , Palash Roy , Md. Abdur Razzaque , Shamsul Huda , Mohammad Mehedi Hassan","doi":"10.1016/j.cor.2024.106877","DOIUrl":"10.1016/j.cor.2024.106877","url":null,"abstract":"<div><div>Online food delivery (OFD) represents a rapidly evolving e-business application that leverages cloud computing data centers, playing a crucial role in meeting the demands of urban lifestyles. With diverse order fulfillment features and increasing expectations for service quality, the task of effectively assigning riders for timely long-distance, cross-regional deliveries presents a significant engineering challenge. Previous studies often relied on traditional rider allocation methods that fail to account for varying capacities, or they utilized non-intelligent systems that did not adequately address fluctuating order demands and service delays. In this study, we introduce a robust Mixed Integer Linear Programming (MILP) optimization framework designed to minimize the total service time and delivery cost for cross-regional orders. This framework divides a large OFD area into multiple regions and utilizes both transfer vehicles and riders to optimize deliveries. To enhance the predictive accuracy of our model, we incorporate advanced machine learning techniques. Specifically, we employ the Long Short-Term Memory (LSTM) model to forecast regional order demands accurately, reflecting the dynamic nature of the marketplace. Additionally, Extreme Gradient Boosting (XGBoost) is tailored to dynamically predict travel times from restaurants to customer locations, facilitating more precise scheduling and resource allocation within the MILP framework. These machine learning techniques significantly bolster the MILP framework by providing detailed, accurate predictions that improve decision-making processes and adaptability to real-time conditions. Acknowledging the complexity of this optimization problem, we further enhance our approach by integrating a meta-heuristic algorithm, Adaptive Large Neighbor Search (ALNS), which efficiently assigns orders to the appropriate transfer vehicles and riders within polynomial time. Our Cross Regional Online Food Delivery (XROFD) system is meticulously designed to optimize both customer satisfaction and rider incentives. Simulation experiments confirm that the XROFD system not only reduces service times and delivery costs but also markedly enhances customer satisfaction and provides superior incentives for riders, outperforming existing state-of-the-art methods.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"173 ","pages":"Article 106877"},"PeriodicalIF":4.1,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-23DOI: 10.1016/j.cor.2024.106876
Ernesto Parra Inza , Nodari Vakhania , José María Sigarreta Almira , Frank Ángel Hernández Mira
A dominating set in a graph is a subset of its vertices such that every its vertex that does not belong to set is adjacent to at least one vertex from set . A set of vertices of graph is a global dominating set if it is a dominating set for both, graph and its complement. The objective is to find a global dominating set with the minimum cardinality. Neither exact nor approximation algorithm existed for the problem known to be -hard. We show that it remains -hard for restricted types of graphs. At the same time, we specify some families of graphs for which the three heuristics, that we propose here, are optimal. Given the complexity status of the problem, our aim was the development of powerful heuristic algorithms that work well in practice for large-scaled instances. To measure the efficiency of our heuristics, we formulated the problem as an integer linear program (ILP) and also we developed an alternative implicit enumeration (IE) algorithm obtaining guaranteed optimal solutions for the existing benchmark instances with up to 8000 vertices. Remarkably, for 56.75% of these instances, at least one of our heuristics also created an optimal solution, where an average absolute error for the remaining instances was a single vertex. The average approximation ratio was 1.005, whereas for the largest benchmark instances with up to 25000 vertices our heuristics delivered solutions in less than 2 min.
图 G 中的支配集 D 是其顶点的一个子集，该子集的每个不属于集合 D 的顶点都至少与来自集合 D 的一个顶点相邻。如果图 G 的顶点集合对图 G 及其补集都是支配集，那么该顶点集合就是全局支配集。全局支配集的目标是找到一个心数最小的全局支配集。对于这个已知的 NP 难问题，既没有精确算法，也没有近似算法。我们证明，对于受限类型的图，该问题仍然是 NP-hard。同时，我们还指出了一些图族，对于这些图族，我们在此提出的三种启发式算法是最优的。考虑到问题的复杂性，我们的目标是开发出强大的启发式算法，并在实践中很好地应用于大规模实例。为了衡量我们的启发式算法的效率，我们将问题表述为整数线性规划（ILP），并开发了另一种隐式枚举（IE）算法，该算法能在顶点多达 8000 个的现有基准实例中获得有保证的最优解。值得注意的是，对于其中 56.75% 的实例，我们的启发式算法中至少有一种也能找到最优解，而其余实例的平均绝对误差仅为一个顶点。平均近似率为 1.005，而对于高达 25000 个顶点的最大基准实例，我们的启发式方法在不到 2 分钟的时间内就给出了解决方案。
{"title":"Algorithms for the global domination problem","authors":"Ernesto Parra Inza , Nodari Vakhania , José María Sigarreta Almira , Frank Ángel Hernández Mira","doi":"10.1016/j.cor.2024.106876","DOIUrl":"10.1016/j.cor.2024.106876","url":null,"abstract":"<div><div>A dominating set <span><math><mi>D</mi></math></span> in a graph <span><math><mi>G</mi></math></span> is a subset of its vertices such that every its vertex that does not belong to set <span><math><mi>D</mi></math></span> is adjacent to at least one vertex from set <span><math><mi>D</mi></math></span>. A set of vertices of graph <span><math><mi>G</mi></math></span> is a global dominating set if it is a dominating set for both, graph <span><math><mi>G</mi></math></span> and its complement. The objective is to find a global dominating set with the minimum cardinality. Neither exact nor approximation algorithm existed for the problem known to be <span><math><mrow><mi>N</mi><mi>P</mi></mrow></math></span>-hard. We show that it remains <span><math><mrow><mi>N</mi><mi>P</mi></mrow></math></span>-hard for restricted types of graphs. At the same time, we specify some families of graphs for which the three heuristics, that we propose here, are optimal. Given the complexity status of the problem, our aim was the development of powerful heuristic algorithms that work well in practice for large-scaled instances. To measure the efficiency of our heuristics, we formulated the problem as an integer linear program (ILP) and also we developed an alternative implicit enumeration (IE) algorithm obtaining guaranteed optimal solutions for the existing benchmark instances with up to 8000 vertices. Remarkably, for 56.75% of these instances, at least one of our heuristics also created an optimal solution, where an average absolute error for the remaining instances was a single vertex. The average approximation ratio was 1.005, whereas for the largest benchmark instances with up to 25000 vertices our heuristics delivered solutions in less than 2 min.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"173 ","pages":"Article 106876"},"PeriodicalIF":4.1,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study mathematical formulations for batch-processing machine scheduling problems (BPMPs), which are the challenging issues in the machine scheduling literature where machines are capable of processing a batch of jobs simultaneously if jobs with non-identical sizes can be packed in a capacitated machine. In this paper, we tackle single- and parallel-machine BPMPs, and other interesting problem variants that aim at minimizing the makespan. We develop novel formulations along with valid inequalities and an algorithm framework that makes use of dual information and bounding techniques to achieve efficiency when instances are intractable. Extensive computational experiments on benchmark instances show that our approaches achieve state-of-the-art results and prove the optimality of intractable instances in the literature.
{"title":"Novel mathematical formulations for parallel-batching processing machine scheduling problems","authors":"Shaoxiang Zheng , Naiming Xie , Qiao Wu , Caijie Liu","doi":"10.1016/j.cor.2024.106859","DOIUrl":"10.1016/j.cor.2024.106859","url":null,"abstract":"<div><div>We study mathematical formulations for batch-processing machine scheduling problems (BPMPs), which are the challenging issues in the machine scheduling literature where machines are capable of processing a batch of jobs simultaneously if jobs with non-identical sizes can be packed in a capacitated machine. In this paper, we tackle single- and parallel-machine BPMPs, and other interesting problem variants that aim at minimizing the makespan. We develop novel formulations along with valid inequalities and an algorithm framework that makes use of dual information and bounding techniques to achieve efficiency when instances are intractable. Extensive computational experiments on benchmark instances show that our approaches achieve state-of-the-art results and prove the optimality of intractable instances in the literature.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"173 ","pages":"Article 106859"},"PeriodicalIF":4.1,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142572184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号