Preventing the generation of inconsistent sets of crisp classification rules
作者:
Highlights:
• The proposed algorithms enhance the generation of sets of classification rules.
• The enhanced process ensures that all the rules are consistent with each other.
• Generating conflict-free rules removes the need for conflict resolution strategies.
• The algorithms can be used with categorical and continuous data.
• The generated rules can be single or multi-label; hierarchical or flat.
摘要
•The proposed algorithms enhance the generation of sets of classification rules.•The enhanced process ensures that all the rules are consistent with each other.•Generating conflict-free rules removes the need for conflict resolution strategies.•The algorithms can be used with categorical and continuous data.•The generated rules can be single or multi-label; hierarchical or flat.
论文关键词:Crisp classification rules,Interpretability,Rule consistency,Constraint handling
论文评审过程:Received 23 August 2019, Revised 29 February 2020, Accepted 29 July 2020, Available online 3 August 2020, Version of Record 17 August 2020.
论文官网地址:https://doi.org/10.1016/j.eswa.2020.113811