文章基本信息

标题：IDS Improved with K-Means Algorithms, Self Organizing Map and Auto Class
本地全文：下载
作者：Vivek Vashishtha ; Durgesh kumar
期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2012
卷号：2
期号：5
出版社：S.S. Mishra
摘要：Intrusi on detecti on system is use to detect sus picious acti vi ties is one for m defence. This paper ai m to bui ld an Intrusion detecti on sys tem that can detect k nown an d unknown network i ntrusions automa tical ly. Under a data mini ng frame work ,the i ds are trained wi th uns upervised le aning alg orit hms , namely the k-means alg orith ms , sel f org anizing map and auto class. Based on thes e uns upervised learning al gori thms , these novel i ds meth ods are proposed and tested . Much hig her detec tion rates are obt aine d wi th reason able true pos iti ve rates, when compare d to the best result obtai ned on the KDD19 99 data set. Moreover, this ids i s mo dularized so as to simplify the incorporatio n of new algorith ms when neces sary .We per form ex peri ments on the tcpdump da ta an d ex trac t appropriate fea ture . Clear disti nctio n between normal a nd a bnormal data is obs erve d when dat a mi ning tec hni ques are applied on t hese features. A series of experi ments h ave been con ducted on KD D19 99 datas et .Different detection methods are trie d to see how they perform i n our i ds. Beari ng KD D 199 9 winner's result a very high detecti on rate has been obtained, alt houg h with a reasona bl e true positi ve r ate.
关键词：KK19 99;weka ; NET la b; K-mean algorith m