Interpretability and accuracy trade-off in the modeling of belief rule-based systems

作者：

Highlights：

• The problem of accuracy-interpretability trade-off in BRB modeling is illustrated.

• The definition and measurement of interpretability in BRB systems are given.

• Propose an optimization framework with accuracy-interpretability trade-off in BRB.

• Prove effectiveness of the proposed framework in classification and prediction.

摘要

•The problem of accuracy-interpretability trade-off in BRB modeling is illustrated.•The definition and measurement of interpretability in BRB systems are given.•Propose an optimization framework with accuracy-interpretability trade-off in BRB.•Prove effectiveness of the proposed framework in classification and prediction.

论文关键词：Interpretability and accuracy trade-off,Belief rule-based systems,Difference evolution algorithm,Single-objective optimization

论文评审过程：Received 29 April 2021, Revised 8 September 2021, Accepted 10 September 2021, Available online 14 September 2021, Version of Record 29 December 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107491