Network optimization is a branch of mathematical programming studying how to plan, optimize, and manage the network to improve its performance, which has been widely investigated and practiced in the fields of transportation, logistics, supply chain, and so on. In recent years, one of the hot research directions on network optimization is to consider the interactions with big data. On one hand, big data is able to achieve the network transparency and visualization, which is helpful on reducing information asymmetry and bullwhip effect. On the other hand, big data can not exclude all uncertainties related to indeterminate quantity or imprecise parameters, such as customer demand, transportation time, etc. In practice, most network optimization approaches are challenged by both transparency and indeterminacy. Therefore, network optimization with big data and uncertain data has received more and more attentions from both academia and industry. A query of Web of Science clearly demonstrates a rapid growth in this research area. The special issue has collected five high-quality papers which develop network optimization models or algorithms in uncertain environments or based on the background of big data. First, as newly strategic analysis of the emerging online marketplace considering risk attitude and channel power, Chen, Hao, and Yan (2020) studied whether the e-tailer and the manufacturer can reach a consensus on introducing the marketplace channel. Second, in order to investigate the multi-level warehouse layout problem with indeterminate information, He et al. (2020) proposed two novel uncertain network optimization models and an effective solution approach, where the indeterminate factors such as the monthly demands and horizontal transportation distances were described by uncertain data. Third, Li et al. (2020) formulated a bi-objective integrated routing optimization approach for rapid and detailed post-disaster needs assessment, and proposed a tabu search algorithm for handling the large-scale optimization problem. The effects of site familiarity and route familiarity were considered in order to achieve a better balance between the contradictory objectives. Fourth, uncertain time series analysis is a new tool to predict future values based on imprecise observations, which are usually described by uncertain data. Lu et al. (2020) proposed two methods for predicting the future value of uncertain time series based on the autoregressive moving average models. Compared with the existing prediction methods, the effectiveness of their work were demonstrated by computational results. Finally, Zhang, Liu, and Liu (2020) proposed least absolute deviations estimations for unknown parameters in uncertain multivariate regression models with imprecise observations. In summary, each article hasmade certain contribution onmodels and algorithms to the development of network optimization with big data and uncertain data.

中文翻译：

---

大数据和不确定数据下的网络优化

网络优化是研究如何规划、优化和管理网络以提高其性能的数学规划的一个分支，在运输、物流、供应链等领域得到了广泛的研究和实践。近年来，网络优化的热点研究方向之一是考虑与大数据的交互。一方面，大数据能够实现网络的透明化和可视化，有助于减少信息不对称和牛鞭效应。另一方面，大数据并不能排除与数量不确定或参数不精确相关的所有不确定性，如客户需求、运输时间等。在实践中，大多数网络优化方法都受到透明度和不确定性的挑战。所以，大数据和不确定数据下的网络优化越来越受到学术界和工业界的关注。Web of Science 的查询清楚地表明该研究领域的快速增长。本期特刊收集了5篇在不确定环境或基于大数据背景下开发网络优化模型或算法的高质量论文。首先，作为考虑风险态度和渠道力量的新兴在线市场的新战略分析，Chen、Hao 和 Yan (2020) 研究了电子零售商和制造商能否就引入市场渠道达成共识。其次，为了研究信息不确定的多层次仓库布局问题，He等人。(2020) 提出了两种新颖的不确定网络优化模型和一种有效的解决方法，其中月需求量、水平运输距离等不确定因素用不确定数据描述。第三，李等人。(2020) 提出了一种用于处理大规模优化问题的禁忌搜索算法。考虑了场地熟悉度和路线熟悉度的影响，以在相互矛盾的目标之间取得更好的平衡。第四，不确定时间序列分析是一种基于不精确观测预测未来值的新工具，通常用不确定数据来描述。卢等人。(2020) 提出了两种基于自回归移动平均模型预测不确定时间序列未来值的方法。与现有的预测方法相比，计算结果证明了其工作的有效性。最后，Zhang、Liu 和 Liu（2020）提出了未知参数的最小绝对偏差估计、具有不精确观测值的不确定多元回归模型。综上所述，每篇文章都在模型和算法上对大数据和不确定数据网络优化的发展做出了一定的贡献。