An Empirical Study of Intrusion Detection System Using Feature Reduction Based on Evolutionary Algorithms and Swarm Intelligence Methods

The use of computer more and more need to be increase the security day by day in a real world, the process of monitoring the computer system in a secure way for an unknown and known attack increase the availability and confidentiality of the system, the system provides the solution of such a problem is known as an Intrusion detection system. Intrusion detection systems play a vital role in network security, its check the system for integrity, confidentiality and availability in an individual manner i.e. Host based intrusion detection system and or in a group i.e. Network based intrusion detection system. Basically an IDS inspects the suspicious behavior of a system in a network may be an attack or misuse of a system. The performance of IDS measures in the detection rate in terms of Precision and recall on the basis of performance parameters such as true positive, false positive, true negative and false negative. In this paper we analyze a existing methods like Neural network techniques, Evolutionary like Genetic algorithm with swarm intelligence based techniques like Particle Swarm optimization for the intrusion detection system to compare the detection rate for the reduced dataset using with experimental result on the basis of KDDCUP dataset to test the performance of various methods and existing algorithm, the dataset include normal and abnormal dataset with the complete features or an attributes. An experimental result shows that the better comparative analysis method provides the results in the form of higher detection rates than other existing techniques.