摘要
传统关联规则挖掘算法主要基于支持度-可信度构架,时空开销的限制使其无法深入挖掘非频繁项集。目前对带类属性的关联分类增量学习研究较少,该文提出一种新的增量式关联分类方法,解决了带类属性数据的增量学习问题,在数据频繁更新时,实现有限时空开销下关联规则的快速提取和维护。实验结果表明,该方法能有效维护并更新关联规则,避免重复学习历史样本,保证分类模型的预测能力。
Traditional associative rule mining algorithm is mostly based on the support-confidence framework, which disable the in-depth study of frequent items for time and space limitations. There is few study of associative classification incremental learning currently. This paper presents a new incremental associative classification method, which can solve the incremental learning problems of data with class attribute, and realize the fast extraction and maintenance of associative rule with limited time and space when the data is updating frequently. Experimental results show that this method can quickly and effectively maintain and update the classification rules, which avoid re-learning the history samples and ensure the predictability of the classification model.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第4期159-161,164,共4页
Computer Engineering
基金
国家自然科学基金资助项目(10771176)
关键词
关联分类规则
增量学习
病毒检测
associative classification rule
incremental learning
malware detection
