摘要
关联分类具有较高的分类精度和较强的适应性。基于闭频繁项集有效压缩事务及FPL(Frequent Pattern List)简单数据结构等方面的优点,提出了一种关联分类器方法。设计了便于分类的FPL变形模式,引入了有效发现闭频繁项集的签名向量合取操作。将闭频繁项集挖掘方法应用于关联分类,提高了关联分类算法的分类效率及准确率。
Associative classification bears fairly high classification accuracy and fairly robust suitability.Based on closed frequent itemset effective compression transaction,FPL(Frequent Pattern List) simple data structure and other merits,the article proposes an associative classifier method,designs a modified FPL pattern to ease classification,and introduces signature vector conjunction operation that effectively discovers closed frequent itemsets. When the closed frequent itemset mining method is applied to associative classification,the classification efficiency and accuracy of the associative classification algorithm is upgraded.
出处
《计算机应用与软件》
CSCD
2011年第6期39-42,共4页
Computer Applications and Software
基金
国家自然科学基金项目(10961017)
关键词
关联分类
分类器
闭频繁项集
向量合取
频繁模式表
Associative classification Classifier Closed frequent itemset Vertex conjunction Frequent Pattern List(FPL)