Differentiating Network Attacks using C4.5 Algorithm with Multiboosting
In most of the security system intrusion detection system is deployed to detect and possibly prevent the attacks. Data mining methods can be used in intrusion detection system. There are several data mining methods like classification tree algorithms (C4.5), Support vector machine and so on. For a detection system to detect an attack efficiently and accurately, it must be able to reduce false positives (identifying an attack when there is no attack in the data stream). In the proposed system multi-boosting technique is used with the C4.5 classification algorithm to reduce false positives and classification errors while classifying the attacks. The attacks that are going to be detected by the system are Denial of Service attacks (DoS), Remote To Local attack (R2L), User To Root attack(U2R).