A joint optimization method on parameter and structure for belief-rule-based systems
作者：
Highlights：
• Propose and justify the new criterion to determine the optimal BRB.
• Propose a heuristic strategy-based algorithm to optimize the structure of BRB.
• Propose a DE-based algorithm to optimize the parameters of BRB.
• Propose a joint optimization method based on new criterion and the proposed algorithms.
• Experiments show the proposed method can achieve the optimal BRB.
摘要
•Propose and justify the new criterion to determine the optimal BRB.•Propose a heuristic strategy-based algorithm to optimize the structure of BRB.•Propose a DE-based algorithm to optimize the parameters of BRB.•Propose a joint optimization method based on new criterion and the proposed algorithms.•Experiments show the proposed method can achieve the optimal BRB.
论文关键词：Belief-rule-based system,Parameter optimization,Structure optimization,Generalization error,Heuristic strategy
论文评审过程：Received 5 April 2017, Revised 1 October 2017, Accepted 29 November 2017, Available online 2 December 2017, Version of Record 17 January 2018.
论文官网地址：https://doi.org/10.1016/j.knosys.2017.11.039
