南京大学学报(自然科学版) ›› 2019, Vol. 55 ›› Issue (4): 564572.doi: 10.13232/j.cnki.jnju.2019.04.006
Jiahui Li1,2,Zhongmei Zhou1,2()
摘要:
基于支持度置信度框架的关联分类算法在生成规则时难以提出大量高质量规则,而且在一些数据集尤其是不平衡数据集上,部分训练实例未被产生的关联规则所覆盖,导致算法的分类准确率不高.基于以上问题提出了改进的关联分类的算法(Improved Algorithm based on Multiple learning and Correlation degree,IAMC).首先,在提取规则时,IAMC对训练集进行多次关联分类学习,尽量多地提出高质量的规则.其次,在生成规则时采用综合考虑了置信度,补类支持度的新度量关联度,以提高生成的规则的质量.最后,在关联分类规则提取后,对利用已有规则无法判断类别的和未被已有规则覆盖的训练实例用决策树方法再次提取规则,并加入到规则集中.实验结果表明,IAMC算法能提出更多高质量的规则,在多个UCI数据集上具有较高的分类准确率.
中图分类号:
1 | YuK,WuX,WeiD,et al. Causal associative classification∥IEEE International Conference on Data Mining Workshops. Brussels,Belgium:IEEE,2012:914-923 . |
2 | LuS H,ChiangD A,KehH C,et al. Chinese text classification by the Na?ve Bayes classifier and the associative classifier with multiple confidence |
threshold values. Knowledge?Based Systems,2010,23(6):598-604. | |
3 | AlwidianJ,HammoB H,ObeidN. WCBA:weighted classification based on association rules algorithm for breast cancer disease. Applied Soft Computing,2017,62:536-549. |
4 | HeH B,GarciaE A. Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering,2009,21(9):1263-1284. |
5 | DengH,RungerG,TuvE,et al. CBC:an asso?ciative classifier with a small number of rules. Decision Support Systems,2014,59:163-170. |
[1] | 许 林,张 巍*,梁小龙,肖 瑞,曹剑秋. 岩土介质孔隙结构参数灰色关联度分析[J]. 南京大学学报(自然科学版), 2018, 54(6): 1105-1113. |
|