The cross-dock door assignment problem (CDAP) is a critical challenge in optimizing cargo transshipment in logistics, where each inbound door serves multiple origins and each outbound door serves multiple destinations, subject to capacity constraints on the volume of goods processed at each door. To address this problem, we propose a novel metaheuristic (LA-HM) that integrates landscape-aware iterated tabu search and large neighborhood search within a unified framework. The proposed algorithm begins with a model-driven procedure to generate a diverse set of initial solutions, which are then refined using an iterated tabu search procedure, enhanced by an innovative landscape-aware detection mechanism. This mechanism dynamically monitors the search landscape at regular intervals to identify plain regions characterized by slow local improvement. Based on the detected plain level, the algorithm adaptively adjusts the interval length in the tabu search and applies targeted perturbations to escape local optima, thereby exploring more promising regions. Additionally, we introduce a reformulation of the original problem, consolidating all trucks at a door into a single “consolidated truck.” This reformulation enables the use of an alternating local search method to further refine solutions efficiently. Evaluated on two widely used data sets, the proposed algorithm sets new best known records for 39 out of 99 benchmark instances, and surpasses the leading algorithms in the literature. Moreover, a comprehensive analysis of the landscape-aware detection mechanism and the large neighborhood search highlights their contributions to solution quality. Notably, the search strategies proposed in this work are generalizable and applicable to a wide range of combinatorial optimization problems, particularly those characterized by flat landscapes where traditional methods often struggle to navigate effectively. History: Accepted by Erwin Pesch, Area Editor for Heuristic Search & Approximation Algorithms. Funding: This work was supported by National Natural Science Foundation of China [Grant 72371200]; Spanish government-Ministerio de Ciencia e Innovación [Grants PID2021-125709OB-C21, PID2024-160226OB-C21]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2025.1233 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2025.1233 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
The software and documentation in this repository are a snapshot of the code archive associated with the paper A preconditioned augmented Lagrangian method for solving semidefinite programming problems by Tianyun Tang and Kim-Chuan Toh.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Exact Algorithms for Two-Dimensional Knapsack Problems: A Unified Framework with New Benchmark Results by Sunkanghong Wang, Roberto Baldacci, Fabio Furini, Lijun Wei, and Qiang Liu. This snapshot corresponds to the version used in the published paper.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Learning to Simulate: Generative Metamodeling via Quantile Regression by L. Jeff Hong, Yanxi Hou, Qingkai Zhang, and Xiaowei Zhang.
We consider the well-known uniform machine scheduling problem [Formula: see text], in which we are given a set of n jobs with processing times [Formula: see text] and a set of m parallel machines, each with a corresponding speed factor [Formula: see text] for [Formula: see text]. The goal is to find an assignment of the jobs to the machines that minimizes the makespan. We propose a general methodology to derive approximation results for different algorithms that share the following high-level structure: a specific procedure is first applied to a list of long jobs, followed by a standard list scheduling approach. This methodology is based on theoretical results we derive for [Formula: see text] that allow us to analyze instances with a limited number of jobs. We evaluate the worst case performance of several algorithms using mathematical programming formulations and derive both existing and new approximation ratios for the problem with up to seven machines. In this context, our approach offers a more flexible tool than traditional analytical proof systems for studying the [Formula: see text] problem and potentially other related problems. History: Accepted by Erwin Pesch, Area Editor for Heuristic Search & Approximation Algorithms. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2025.1129 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2025.1129 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
With the rapid rise of generative artificial intelligence (AI), the responsible development and governance of AI and data science have become central concerns in both academic research and practice. As generative AI systems become increasingly embedded in daily life and play a growing role in decision making, it is essential that they operate in ethical, transparent, accountable, and socially responsible ways. In this editorial, we examine how responsible AI and data science can create meaningful societal impact across six substantive areas: judicial systems, education, communication, healthcare, bias and fairness, and interpretability. Rather than treating principles such as fairness, bias mitigation, transparency, accountability, privacy protection, robustness, interpretability, and social impact as separate organizational pillars, we view them as cross-cutting design principles that arise across these domains and methodological areas. We discuss how these principles can inform the design, deployment, and governance of AI systems that address complex societal challenges, safeguarding human values and ethical standards. Finally, we outline future research directions by contrasting pregenerative AI priorities with the emerging challenges of the postgenerative AI era. In doing so, we identify computational, methodological, and optimization frameworks that can support the responsible development and deployment of generative AI systems for meaningful societal benefit.
In this paper, we introduce the novel concept of clustered interpretability to enhance the transparency and explainability of artificial intelligence models in data sets with heterogeneous feature-outcome relationships. Traditional interpretability techniques often focus on global or local explanations, providing either an overview of the model’s behavior across the data set or explanations for individual predictions. However, these approaches can fall short when modeling clustered data sets, in which interpretable feature-outcome relationships across subgroups are desirable. Furthermore, existing research on interpretable clustering generally focuses on the interpretability of cluster assignment decisions in unsupervised settings rather than the interpretability of feature effects in supervised settings. To address this gap, our framework proposes a novel metric—the separation-disparity index—that prioritizes between-cluster variance while minimizing unnecessary heterogeneity in the clusters’ feature-outcome relationships. We use this metric alongside particle swarm optimization (PSO) to learn interpretable cluster solutions, incorporating novel modifications into the PSO algorithm to encourage high-quality solutions. We then demonstrate the value of our framework on synthetic data and a real-world healthcare data set of Parkinson’s disease patients. History: This paper has been accepted by Kaushik Dutta for the Special Issue on Responsible AI and Data Science for Social Good. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.0657 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.0657 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
We consider the distributed optimization problem with data dispersed across multiple workers under the orchestration of a parameter server. In distributed environments, variations in computation speeds and network conditions across workers often lead to significant idle times in synchronous training. Although asynchronous training has been widely explored to reduce the synchronization overhead, existing methods either assume bounded dissimilarity among workers' local data, which hampers performance under high data heterogeneity, or rely on worker scheduling strategies that limit system asynchrony. This work proposes the dual-delayed stochastic gradient descent (DuDe-SGD) algorithm to overcome the above limitations. Through a server-side buffer architecture, DuDe-SGD makes use of stale stochastic gradients from all workers to neutralize the effects of data heterogeneity while maintaining full asynchrony and per-iteration computation cost on par with traditional asynchronous stochastic gradient descent (SGD) algorithms. Our analysis demonstrates that DuDe-SGD achieves a comparable convergence rate for smooth nonconvex problems as state-of-the-art asynchronous SGD algorithms, even with arbitrarily heterogeneous data without adopting any worker scheduling schemes. Numerical experiments demonstrate the favorable performance of DuDe-SGD compared with existing synchronous and asynchronous SGD-based algorithms, especially in scenarios with highly heterogeneous data.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Integrating Mental Health and Juvenile Justice Outcomes: A Case Study of Model-Agnostic Interpretable Machine Learning by Monica Chiarini Tremblay, Arturo Castellanos, Rajiv Kohli, Erin Espinosa, and Thomas Roderick.
The software and data in this repository are a snapshot of the software and data that were used in the research reported in the paper On the approximation of separable non-convex optimization programs to an arbitrary numerical tolerance by Claudio Contardo and Sandra U. Ngueveu.
We study a variant of online assortment optimization under multinomial logit (MNL)-Bandit feedback where, in addition to maximizing the total revenue, the seller must also take fairness into account. We consider the fairness constraint that the seller must offer each product for a predefined fraction of time at any time, giving every product a fair chance for exposure. We first propose two algorithms that satisfy fairness constraints with high probability: a maximum likelihood estimation (MLE)–based algorithm and an upper confidence bound (UCB)–based algorithm, both of which rely on solving linear programs (LPs). These algorithms achieve regret bounds of [Formula: see text] and [Formula: see text], respectively. Next, we propose an MLE-based algorithm and a UCB-based algorithm that both satisfy the fairness constraints and avoid solving complex LPs by calculating efficiently solvable assortment optimization problems, with the UCB-based algorithm achieving a [Formula: see text] regret under the separation assumption for the revenue and utility parameters. Finally, we empirically validate the theoretical results on real data. History: Accepted by Andrea Lodi, Area Editor for Design & Analysis of Algorithms – Discrete. Funding: This work was supported by the Shanghai Education Development Foundation [Grant 23CGA02] and the National Natural Science Foundation of China [Grant 12301376]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2025.1364 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2025.1364 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper On the Use of Regular Languages to Model Personnel Scheduling Problems by G. Ghienne, O. Bellenguez, G. Massonnet and M.I. Restrepo.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Local Search for Integer Quadratic Programming by Xiang He, Peng Lin, Tao Jiang and Shaowei Cai.
The software and replication materials in this repository are a snapshot of the software and data that were used in the research reported on in the paper Fast Filter Pruning Method for Neural Network Compression by D. Wang, W. Chen, C. Hua, W. Zhou, and X. Lin.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper An Efficient Solver for Integral Flows in Decision Hypergraphs with Applications to Cutting Problems by A. Léonard and F. Clautiaux.
The software and data in this repository are a snapshot of the software and data that were used in the research reported in the paper Contextual Optimizer through Neighborhood Estimation for Prescriptive Analytics by Xiao Jin, Yichi Shen, and Loo Hay Lee.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper An Interpretable Preference Learning Model Admitting Dynamic and Context-Dependent Preferences by Zice Ru, Jiapeng Liu, Miłosz Kadziński, Xiuwu Liao, and Xinlong Li.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Semantic Aggregated Adversarial Training Framework for Hate Speech Detection by Xingwei Zhang, Hu Tian, Xiaolong Zheng, Jing Peng and Daniel Dajun Zeng.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper Contextual Stochastic Vehicle Routing with Time Windows by Breno Serrano, Alexandre M. Florio, Stefan Minner, Maximilian Schiffer, and Thibaut Vidal. This snapshot corresponds to the version used in the published paper.
The software and data in this repository are a snapshot of the software and data that were used in the research reported on in the paper MNL-Bandit with Fairness Constraints by Zikun Zhang and Guanhua Fang.
