摘要
针对如何提高决策林的分类精度问题,提出一种基于粗糙集约简构建决策林的技术,包括基于逐次数据约简构建粗糙决策林和基于遗传算法构建粗糙决策林。对3个UCI数据集的验证表明,基于遗传算法构建的粗糙决策林获得了更好的分类效果。
This paper proposes a technique to construct decision forests based on rough set reduction to enhance the classification performance of decision forests. It includes two methods： one is based on sequentially data reduction to construct rough decision forests, the other is based on genetic algorithm to construct rough decision forests. Experimental results in three data sets of UCI show that the rough decision forests constructed by genetic algorithm get better classification performances.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第15期193-194,197,共3页
Computer Engineering
基金
东北林业大学青年科研基金资助项目(07024)
关键词
决策林
粗糙集
约简
遗传算法
decision forests
rough set
reduction
genetic algorithm
