A BAYESIAN NETWORK APPROACH TO IMPROVE CHANGE PROPAGATION ANALYSIS
Year: 2015
Editor: Christian Weber, Stephan Husung, Gaetano Cascini, Marco Cantamessa, Dorian Marjanovic, Frederico Rotini
Author: Lee, Jihwan; Hong, Yoo S.
Series: ICED
Institution: Seoul National University, Korea, Republic of (South Korea)
Section: Design Methods and Tools - part 1
Page(s): 253-264
ISBN: 978-1-904670-68-1
ISSN: 2220-4334
Abstract
Predicting and analysis of change propagation is one of the important issues in engineering change management. In this paper, we illustrate how the Bayesian network (BN), which is emerging tool for a wide range of risk management, can be used in change propagation analysis. The paper compares BN with most popular tool for change propagation analysis, namely CPM. Firstly, we show that CPM can be directly converted into an equivalent BN. In addition to this, we also show that BN has significant advantages over CPM at both modeling and analysis level. In the model-ing level, various assumption over CPM can be relaxed and various kinds of modeling exten-sion can be accommodated with BN. At the analysis level, BN s ability to performing probabilistic inference provides a user with another interesting measures, which cannot be obtained with CPM. Moreover, BN provides robust framework for learning change propagation probabilities from empirical data. Our method can enhance the capability of engineering change management throughout entire product life-cycle.
Keywords: Change Management, Change Propagation, Bayesian Network, Design Management, Design Informatics