摘要
在数据挖掘过程中,有很多挖掘算法试图使离群点的影响最小化,甚至是排除它们,然而这样可能丢失一些重要的信息。如今,在欺诈检测、网络入侵检测、故障诊断等问题中,离群点挖掘得到了越来越多的应用,离群点的发掘成为一个热门研究问题。I-Miner是一个企业级的数据挖掘工具,在本文中利用I-Miner软件对数据进行预处理,并用通过S语言拓展软件功能,编写了3种离群点算法并使用多个数据测试,对结果进行分析和对比研究。
Most algorithms try to minimize or eliminate the effects of the outliers in data mining procedures, but important information may be dropped in these algorithms. Nowadays, the outliers mining is widely used in fraud detection, network intrusion detection and fault diagnosis, research on outlier mining has become an important research field. I-Miner is a data mining tool with enterprise-strength. Three outliers mining algorithms are presented by data pretreatment with I-Miner and function expansion with the S programming language. Running data tests have done with the three algorithms, and analysis and comparison of the results are researched.
出处
《软件》
2011年第11期25-28,31,共5页
Software
