The Research Of Intrusion Detection Model Based On FSVM And Rough Set

Intrusion detection technology can deal with some troublesome problems which the static protection can’t deal with.Intrusion detection based on rough set(RS) and fuzzy support vector machine(FSVM) can get a good performance.RS can effectively analyze and deal with kinds of incomplete information,and find implicit information from it.FSVM can effectively solve the problems of small sample size,high dimension,nonlinear,over learning,local optimal solution and so on.In this paper,the intrusion detection model which combines the advantages of RS and FSVM is introduced.It has been proved efficiently and precisely.Based on the study of intrusion detection system(IDS),RS and FSVM, my research work and innovation as follows:1. Proposing an IDS model based on RS and FSVM.Firstly, The model use RS to reduce the features,then classify the data according to the decision feature,FSVM can deal with the reduced data quickly.Then RS send the legal data which have been classfied by RS to FSVM,and filter again.Double filter make IDS accurately.2. Introducing the definition of feature contribution weighted,it combines advantages of feature contributions which are from information entropy algorithm and feature reduction algorithm,and overcomes the incomplete of feature contribution,we can get more accurate feature contributions.3. Combining with the weighted feature contribution and the fuzzy membership function based on distance between sample point and center point, proposing a new membership function based on feature correlativity,it can distinguish support vector and outlier noise efficiently.