出版社：International Institute for Science, Technology Education
摘要：In this paper, we present a novel method for the detection of outlier in intrusion detection system. The proposed detection algorithm, are called hybrid algorithm. It is combination of two algorithm k-mean and boxplot. Experimental results demonstrate to be superior to existing SCF algorithm. One of the most common problems in existing SCF technique detection techniques is that such as ignoring dependency among categorical variables, handling data streams and mixed data sets. Moreover, identifying number of outliers in advance is an impractical issue in the SCF algorithm and other outlier identification techniques. This paper investigates the performances of boxplot-mean method for detecting different types of abnormal data.
其他摘要：In this paper, we present a novel method for the detection of outlier in intrusion detection system. The proposed detection algorithm, are called hybrid algorithm. It is combination of two algorithm k-mean and boxplot. Experimental results demonstrate to be superior to existing SCF algorithm. One of the most common problems in existing SCF technique detection techniques is that such as ignoring dependency among categorical variables, handling data streams and mixed data sets. Moreover, identifying number of outliers in advance is an impractical issue in the SCF algorithm and other outlier identification techniques. This paper investigates the performances of boxplot-mean method for detecting different types of abnormal data. Keywords : Outlier detection techniques, clustering, scf, genetic and boxplotmean technique.
关键词：Outlier detection techniques; clustering; scf; genetic and boxplotmean technique.
