分享兴趣，传播快乐，

增长见闻，留下美好。

亲爱的您，

这里是LearingYard学苑！

今天小编为大家带来

《复杂供应网络系统风险评估》精读。

欢迎您的用心访问！

本期推文阅读时长大约5分钟，请您耐心阅读。

Share interests, spread happiness,

increase knowledge, and leave beauty behind.

Dear you,

this is LearningYard Academy!

Today, the editor brings you intensive reading of "Systemic Risk Assessment in Complex Supply Networks".

Welcome your visit!

The reading time of this tweet is about 5 minutes, please read it with patience.

小编将从思维导图、精读内容、知识补充三个板块为大家带来《Systemic Risk Assessment in Complex Supply Networks》案例研究1。

I will bring you Case Study 1 of Systemic Risk Assessment in Complex Supply Networks from three sections: Mind Map, Intensive Reading Content, and Knowledge Supplement.

01

思维导图

02

精读内容

文章使用本田-讴歌的物料流和合同关系供应网络作为案例，来说明系统风险概念如何应用于供应网络。

The article uses Honda-Acura's material flow and contractual relationship supply network as a case study to illustrate how the concept of system risk can be applied to supply networks.

一、相关性

1. Relevance

在计算了前两周介绍的每个度量之后，对每个系数进行Pearson相关系数和99.95%置信区间的显著性t检验。粗体数字表示在99.95%置信区间通过t检验的相关性。

After calculating each metric introduced in the first two weeks, Pearson correlation coefficients and t-tests for significance at 99.95% confidence intervals were performed for each coefficient. Bold numbers indicate correlations that passed the t-test at the 99.95% confidence interval.

出现了三种类型的度量:度类度量、接近性类度量和中介性类度量。度类度量包括度中心性、特征向量中心性、枢纽和权威中心性。接近性类度量包括近似度、中心性和径向性。中介性类度量由中介中心性组成。文章分析了组内这些指标的相关性，并决定哪些指标将应用于本田讴歌的案例研究。

Three types of metrics emerged: degree class metrics, proximity class metrics, and mediator class metrics. The degree class metrics include degree centrality, feature vector centrality, pivot and authority centrality. Proximity class metrics include proximity, centrality, and radiality. The intermediation class metric consists of intermediation centrality. The article analyzes the relevance of these metrics within the group and determines which metrics will be applied to the Honda Acura case study.

案例研究中没有考虑度中心性，因为它只部分地解释了网络拓扑。由于特征向量中心性和权威中心性与枢纽中心性不相关，文章在案例研究中考虑了这三者。接近中心性和径向中心性是高度相关的，因此选择其中一个应用在本田-讴歌网络中。之所以选择接近中心性，是因为它在风险评估文献中有很多应用。相关性研究的结果是针对本田-讴歌网络的，因为不同的供应网络可能具有不同的拓扑特征，因此得出的指标可能具有不同的相关性。

Degree centrality is not considered in the case study because it only partially explains the network topology. Since feature vector centrality and authority centrality are not correlated with hub centrality, the article considers all three in the case study. Proximity centrality and radial centrality are highly correlated, so one of them is chosen to be applied in the Honda-Acura network. Proximity centrality was chosen because it has many applications in the risk assessment literature. The results of the correlation study are for the Honda-Acura network because different supply networks may have different topological characteristics and therefore the resulting metrics may have different correlations.

二、实验描述

2. The description of the experiment

文章为未加权和加权网络的每个指标进行了两个实验，强调如何结合领域相关信息可以增强风险分析。为了衡量系统风险，使用了两种不同的风险类型:节点风险和链接风险。对于节点风险，使用世界风险指数(WRI)，对于链接风险，使用两个地理位置之间的距离。在Katz中心性(一种特征向量中心性)研究中，将WRI分配给节点，以识别无向契约供应网络中的风险扩散。在枢纽和权威研究中，将地理位置之间的距离分配给链接，以评估有向材料网络中的供应商和客户风险敞口，在接近性研究中，以评估无向材料网络中中断蔓延的速度。

The article conducts two experiments for each metric of the unweighted and weighted network to highlight how combining domain-related information can enhance risk analysis. To measure systemic risk, two different types of risk are used:node risk and link risk. For node risk, the World Risk Index (WRI) was used, and for link risk, the distance between two geographic locations was used. In the Katz centrality (a type of eigenvector centrality) study, the WRI is assigned to nodes to identify the spread of risk in the undirected contract supply network. In hub and authority studies, we assign distances between geographic locations to links to assess supplier and customer exposures in directed material networks, and in proximity studies to assess the rate of disruption spillover in undirected material networks.

采用距离的倒数研究中介中心性，识别关键中介。实验共八项，实验总结见下表。

The inverse of the distance was used to study mediated centrality and identify key mediators. A total of eight experiments were conducted, and the experiments are summarized in the following table.

Marklines汽车数据库被用于提取公司总部位置，这有助于评估节点和链路权重。从在线数据库中检索到的值和所找到的地理位置应仅作为指导，而不是当前本田-讴歌供应链的实际状态。

The Marklines Automotive database was used to extract the corporate headquarters location, which helped to evaluate node and link weights. The values retrieved from the online database and the geographic locations found should be used as a guide only and not the actual state of the current Honda-Acura supply chain.

由于Intek、JFC、J1、J2、J3等5家企业为了保护企业的身份而更改了名称，因此资料不完整。因此，Intek和JFC被假定为位于中国的供应商。J1、J2和J3是位于日本的供应商，但文章无法在Marklines数据库中找到其他信息，因此假设总部位于东京。文章使用Python 3.4的NetworkX库进行计算。

The information is incomplete because five companies, Intek, JFC, J1, J2, and J3, have changed their names to protect the identity of the companies. Therefore, Intek and JFC are assumed to be suppliers located in China. J1, J2, and J3 are suppliers located in Japan, but the article could not find additional information in the Marklines database, so the headquarters are assumed to be located in Tokyo. The article uses the NetworkX library of Python 3.4 for the calculations.

03

知识补充

近年来，对全球范围内自然灾害相关风险的信息和理解的需求不断增加。这些信息对从事减少灾害风险、空间规划和（再）保险领域的利益相关者至关重要。World Risk Index提供了一种新的方法来评估国家范围内的风险和对自然灾害的脆弱性，并允许在全球范围内对各国进行比较。世界风险指数的概念是指自然灾害和灾害风险界对风险的理解，其中灾害风险被定义为物理灾害和暴露元素的脆弱性相互作用的产物。

The demand for information and understanding on natural hazard related risk on a global scale has grown in recent years. Such information is crucial for stakeholders who are working in the field of disaster risk reduction, spatial planning and (re-)insurance. The World Risk Index provides a new approach to assess risk and vulnerability towards natural hazards on country scale and allows the comparison of countries at global scale.The concept of the World Risk Index is referring to the understanding of risk within the natural hazards and disaster risk community, where disaster risk is defined as the product of the interaction of physical hazards and the vulnerabilities of exposed elements .

世界风险指数的最终结构成分和指标:

Final Structural components and indicators of the World Risk Index:

今天的分享就到这里了。

如果您对今天的文章有独特的想法,

欢迎给我们留言，

让我们相约明天。

祝您今天过得开心快乐！

That's all for today's sharing.

If you have a unique idea for today’s article,

please leave us a message,

and let us meet tomorrow.

I wish you a nice day !

参考资料：Deepl、CSDN、百度

参考文献：

[1]Ledwoch A, Brintrup A, Mehnen J, et al. Systemic risk assessment in complex supply networks [J]. IEEE Systems Journal, 2016, 12(2): 1826-1837.

本文由LearningYard学苑原创，如有侵权请在后台留言！

文字|Zheng

排版|Zheng

审核|wang