Implementation of network intrusion detection system using variant of decision tree algorithm

As the need of internet is increasing day by day, the significance of security is also increasing. The enormous usage of internet has greatly affected the security of the system. Hackers do monitor the system minutely or keenly, therefore the security of the network is under observation. A conventional intrusion detection technology indicates more limitation like low detection rate, high false alarm rate and so on. Performance of the classifier is an essential concern in terms of its effectiveness; also number of feature to be examined by the IDS should be improved. In our work, we have proposed two techniques, C4.5 Decision tree algorithm and C4.5 Decision tree with Pruning, using feature selection. In C4.5 Decision tree with pruning we have considered only discrete value attributes for classification. We have used KDDCup'99 and NSL_KDD dataset to train and test the classifier. The Experimental Result shows that, C4.5 decision tree with pruning approach is giving better results with all most 98% of accuracy.