CAJ | 학술논문

不平衡数据集广泛存在于入侵检测、医疗诊断等领域，研究不平衡数据集的分类问题具有极其重要的应用价值。不平衡数据集分类中，具有重要价值的正类分类存在误差较大的问题。为解决这个问题，本文将聚类算法与局部支持向量机进行融合，提出了一种解决不平衡数据集分类的思路。为验证本文提出算法的有效性，本文使用UCI国际标准数据库进行实验验证，实验结果表明，本文算法能够有效提高不平衡数据集的分类精度。
불평형수거집엄범존재우입침검측、의료진단등영역，연구불평형수거집적분류문제구유겁기중요적응용개치。불평형수거집분류중，구유중요개치적정류분류존재오차교대적문제。위해결저개문제，본문장취류산법여국부지지향량궤진행융합，제출료일충해결불평형수거집분류적사로。위험증본문제출산법적유효성，본문사용UCI국제표준수거고진행실험험증，실험결과표명，본문산법능구유효제고불평형수거집적분류정도。
Unbalanced data sets are widely exists in areas such as intrusion detection, medical diagnosis. The research about unbalanced data sets classification is of great practical value.To solve the problem of positive class in Unbalanced data sets having big cassification error, this paper combine clustering algorithm with local support vector machine and proposes a new solution to the classification problem of unbalanced data sets . In order to verify the validity of the method, this paper did tests on UCI datasets. The experimental results show that the algorithm of this paper can improve the classification accuracy of unbalanced data sets.