Product quality reliability analysis based on rough Bayesian network

Wanjuan Zhang, Xiaodan Wang, Diego Cabrera, Yun Bai

Research output: Contribution to journalArticlepeer-review

4 Scopus citations


Simultaneous quality reliability analysis can detect the weak links in production process as early as possible, which can significantly improve product reliability. Aiming at the reliability in product quality, a model based on rough set and Bayesian network (RS-BN) is proposed in this paper. Simplify expert knowledge and reduce product quality factors using rough set theory, and the minimal product quality rules can be obtained. Then the Bayesian network is constructed and trained by the minimum rules. Based on the minimal rules, the complexity of Bayesian network structure and the difficulties of product reliability analysis are largely decreased. To verify the performance of the proposed RS-BN model, a competition dataset is utilized and four evaluation indicators are investigated, i.e., accuracy, F1-score, recall, and precision. Experimental results indicated that the proposed model is superior to the other three comparative models.

Original languageEnglish
Pages (from-to)37-47
Number of pages11
JournalInternational Journal of Performability Engineering
Issue number1
StatePublished - 1 Jan 2020

Bibliographical note

Funding Information:
This research is supported in part by the National Natural Science Foundation of China (71801044), the National Key Research & Development Program of China (2016YFE0132200), the MOST Science and Technology Partnership Program (KY201802006), the Natural Science Foundation of Chongqing (cstc2018jcyjAX0436), and the Graduate Student Innovation Research Project of Chongqing Technology and Business University (yjscxx2018-060-16).

Publisher Copyright:
© 2020 Totem Publisher, Inc.


  • Bayesian network
  • Minimum rules
  • Product quality
  • Reliability analysis
  • Rough set


Dive into the research topics of 'Product quality reliability analysis based on rough Bayesian network'. Together they form a unique fingerprint.

Cite this