Intrusion Detection System by using SVM and Ant Colony

Abstract

In today’s world, it is very difficult to protect our system from different types of attacks or viruses . Intrusion Detection System (IDS) are basically designed to defend computer system from various types of attacks. In this paper a new approach is proposed for network intrusion detection. This new approach will make the combination of SVM algorithm with Self-Organized Ant Colony Network (CSOACN) algorithm to develop new IDS. It will take advantages of both while avoiding their limitations. The basic task of this approach is to classify network packet as normal or abnormal while minimizing misclassi?cation. For experimental setup we had used standard benchmark NSLKDD dataset which is advanced version of KDDCUP99 dataset. The system shows best result as we compare it with different types of IDS which was developed by using different algorithm.