摘要
将互信息引进模糊决策树,用于确定决策树的候选分类属性,进而构建模糊决策树.通过增量学习来修正决策树分类模型,以修正分类效果,并用实验验证了该方法的有效性.
The fuzzy decision tree used to determine a candidate tree classification attributes based on mutual infor- mation has bcen introduced, and then construct fuzzy decision trees. It corrects classification model to achieve better classification results through incremental learning. Finally, experimentally validated.
出处
《江西师范大学学报（自然科学版）》
CAS
北大核心
2014年第1期89-94,共6页
Journal of Jiangxi Normal University(Natural Science Edition)
基金
教育部重点实验室基金(110411)
广东省自然科学基金(10451009001004804
9151009001000007)
广东省科技计划(2012B091000173)资助项目
关键词
模糊决策树
信息熵
互信息
增量学习
fuzzy decision tree
conditional information entropy
mutual information
ensemble learning
incremental learning
