Intrusion detection system based on bagging with support vector machine

<p>Due to their rapid spread, computer worms perform harmful tasks in networks, posing a security risk; however, existing worm detection algorithms continue struggle achieve good performance and the reasons for that are: First, large amount of irrelevant data affects classification accuracy. Second, individual classifiers do not detect all types effectively. Third, many systems are based on outdated data, making them unsuitable new species. The goal study is use mining network because they have high ability accurately. proposal UNSW NB15 dataset uses support vector machine train test ensemble bagging algorithm. To various efficiently, contribution suggests combining correlation Chi2 feature selection method called Chi2-Corr select relevant features using (SVM) system achieved accuracy reaching 0.998 with Chi2-Corr, 0.989, 0.992 chi-square separately.</p>