摘要
孤立点检测是数据挖掘中一个重要方面,用来发现“小的模式”(相对于聚类),即数据集中显著不同于其他数据的对象.在以往的数据挖掘应用中,孤立点经常被当作“噪声”而被剔除.为更好地提高入侵检测系统中的实时性和准确性,提出了新的解决方案.
Outlier detection is an important aspect in data mining, which is used to find ＂mini pattern＂ （compared with clustering）, the object, which is different from others in data set. Outlier is usually regarded as ＂noise＂ and deleted in data mining application in the past. In order to improve the real time and accuracy of the intrusion detection system, a new solution is presented.
关键词
数据挖掘
孤立点挖掘
入侵检测
data mining
outlier mining
intrusion detection
