期刊名称：International Journal of Computer Information Systems and Industrial Management Applications
印刷版ISSN：2150-7988
电子版ISSN：2150-7988
出版年度：2011
卷号：3
页码：61-71
出版社：Machine Intelligence Research Labs (MIR Labs)
摘要：In data analysis and decision making we need frequentlyto judge whether the observed data items are normal orabnormal. This happens in banking, credit card use, diagnosingpatient health state, fault detection in an engine or devicelike an off-shore oil platform or gearbox in an airplane motor.Sometimes the normal cases are boring and only the abnormalcases are of interest. In practice, it happens quite frequentlythat the normal state has a good representation, however theabnormal cases are rare and the abnormal class is ill-defined;in such a case we have to judge on the abnormality using informationfrom the normal class only. The problem is called ’oneclassclassification’ (OCC). The paper gives a survey of methodsfor performing the OCC. We show also an example: how to detecta masquerader (non-legitimate user) in a computer system– when observing a sequence of commands several thousandslong.
