貝式網路對供應鏈中斷風險之分析

A Bayesian Network Approach for Analyzing Supply Chain Disruption Risk

莊孟蓓(Meng-Pei Chuang)

指導教授 ： 陳以明

元智大學/工程學院/工業工程與管理學系/碩士(2012年)

摘要

有鑑於過去探討供應鏈管理（Supply Chain Management）都著重在報酬利潤，但在整個供應鏈的體系中，上下游的通路關係是相當重要的，只要其中一環發生問題都會造成嚴重損失，因此如何避免供應鏈間的中斷問題就成為管理者的主要目標。一個好的供應鏈中斷風險分析系統可以提出建議和發展來幫助企業在高風險的經營環境下可以不易受到影響，維持公司正常營運。使用此系統執行分析供應鏈，不僅能辨識中斷風險，還能夠更加了解供應鏈中的潛在曝險，在面對供應鏈的中斷風險時做更合理的決策。使用貝式網路（Bayesian Networks）自不確定的訊息中分析供應鏈中斷之風險，而貝式網路是基於邏輯推理及歷史訊息，對供應鏈系統的不斷更新訊息，來計算其新的中斷風險。最後本研究透過建置一個供應鏈中斷風險分析之貝式網路，做供應鏈風險的預測分析，以避免供應鏈中斷發生的風險。

關鍵字

供應鏈管理 ； 風險分析 ； 貝式網路

並列摘要

In view of the past explore the Supply Chain Management (SCM) focus on reward profits. However upstream and downstream pathway relationship is very important to avoid serious losses in the supply chain system. Therefore, how to avoid supply chain disruptions is the main objective of the managers. A good supply chain disruption risk analysis system can make suggestions and development to help companies in high-risk business environment can be less affected, to maintain the companies in normal operation. Use this system to analysis the supply chain can not only able to identify the risk of disruptions, but also better understanding of the potential exposure s to make more rational decision-making in the face of the risk of supply chain disruptions. Bayesian networks reasonings assess under uncertainty in casual structures to analysis the supply chain disruption risk. We also combine new coming information and historical information to calculate the new disruption risk. This study, finally, provides supply chain disruption risk analysis system which used Bayesian networks to predict analysis and avoid the risk of supply chain disruptions.

並列關鍵字

Supply Chain Management ； Risk analysis ； Bayesian network

參考文獻

[1] Jenkins B. D.,"Security Risk Analysis and Management ", 1998.

[2] Cristina Elena Manfredotti, "Modeling and Inference with Relational Dynamic Bayesian Networks", University of Milan Bicocca, 2009.

[4]Harri Lahdesmaki, "Relationships between probabilistic Boolean networks and dynamic Bayesian networks as models of gene regulatory networks, " Signal Processing, 86(4), pp. 814-834, April 2006.

[6] JiaweiHan, Micheline Kamber, Data Mining : Concepts and Techniques, Morgan KaufmannPub,2006.

[7] Mark Treleven, Sharow Bergman Schweikhart, "A risk/benefit analysis of sourcing strategies: single vs. multiple sourcing", Journal of Operations Management, 7 (4), pp.93–114 , 1988.