A Hybrid Approach to improve the Anomaly Detection Rate Using Data Mining Techniques

Abstract

An Intrusion Detection System is a device or software application that monitors events occurring on the network and analyzes it for any kind of malicious activity that violates computer security policy. With an increase in dependency rate on the internet, there is a significant increase in the number of internet attacks as well. The challenges arise towards the network security due to the introduction of new methods of attacks. To identify these attacks, a new hybrid approach using data mining based on C4.5 and Meta algorithm is proposed.This approach provides a classifier which improves the overall accuracy of detection. Various data mining techniques have been developed for detecting intrusion. For detection of anomalies a hybrid technique based on C4.5 and meta-algorithm is proposed that provides better accuracy and reduces the problem of high false alarm ratio. The comparison of the proposed approach is made with other data mining techniques. With this proposed approach detection rate is improved considerably. The experimentation is implemented in WEKA tool using KDD Cup 1999 dataset