摘要
在不平衡数据集中,多数类不一定是最优的,这一问题将会影响决策树的生成效果和分类预测的准确性,提出类置信度比例决策树算法,这种算法对类的大小不敏感.通过实验验证,这种算法比传统的决策树算法更具有优越性.
The majority class in the imbalanced data sets is not necessarily the best, which will influence not only the generating effect of decision tree but also the accuracy of the final classification prediction. Therefore, Class Confidence Proportion Decision Tree is proposed. This new algorithm is insensitive to the size of the classes. It is proved that this algorithm is better than the traditional one through extensive experiments.
出处
《河南师范大学学报（自然科学版）》
CAS
北大核心
2013年第2期154-157,共4页
Journal of Henan Normal University(Natural Science Edition)
基金
河南省科技攻关项目(102102210175)
河南省青年骨干教师资助项目(2010GGJS-068)
关键词
不平衡数据集
类置信度比例
决策树
算法
imbalanced data sets
Class Confidence Proportion
decision Tree
algorithm
