Evolving multi-label classification rules by exploiting high-order label correlations
Neurocomputing(2020)
摘要
•MLR is a multi-label classification method that extends UCS through the LP technique.•MLR exploits high-order label correlations and addresses the incompleteness of LP.•The theoretical analysis provides approximate bounds for the expected rule fitness.
更多查看译文
关键词
Multi-label classification,High-order label correlations,Label powerset,Learning classifier systems,Genetic algorithms
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要