Recent advances in the field of Artificial Intelligence has urged many security experts to follow this novel area. Extensive work has already been done, and much more efforts are being put to use the techniques of AI to improve the real time security. With lot of algorithms and techniques already in the market, it has become quite difficult to choose a particular algorithm, judging the fact that none proves to be accurate and efficient than the others. During the detailed study of major machine learning algorithms, we found certain algorithms more accurate than the other. In this paper, we have chosen two different techniques of differentiating the normal traffic from the intrusions. The results show J48 more efficient and accurate than the Naïve Bayes algorithm.