摘要
显露模式(EP)是支持度从一个数据集到另一个数据集发生显著变化的项集.EP具有很强的区分能力,可以建立很好的分类器.文中采用基于EP的分类算法CEEP建立基分类器,结合组合学习分类方法AdaBoost算法的思想,提出了一种新的分类算法A-E算法.算法使用加权样本建立基分类器,并根据分类结果改变样本权值,同时应用分类误差计算基分类器权重.最终,算法按权重组合每个分类器的分类结果.在UCI机器学习数据库的9个基准数据集上的实验表明,A-E算法都能有效地减低泛化误差,并具有较高的分类准确率.
Emerging pattem（EP） are itemsets whose supports change significantly from one data class to another. EP are very strong at differentiating samples between classes, so they are useful for constructing accurate classifiers. This work proposes a novel EP-based classification method（A-E）, which classification by classifiers with weights. The algorithm construct the basic classifier with the weighted sample, meanwhile calculate the classifier weights with the classification error. In the test, A-E can aggregate differentiating powers of classifiers. Our experiment study carried on 9 benchmark datasets from the UCI Machine Learning, A-E algorithm can effectivdy reduce the generalization error, and has high classification accuracy.
出处
《微电子学与计算机》
CSCD
北大核心
2008年第8期229-231,235,共4页
Microelectronics & Computer
