Improving the efficiency of an Intrusion Detection System using Random Forest and K-Means algorithms

Intrusion Detection System (IDS) is a system that provides a layer of security to an organization’s networks. In today’s world, the number of devices that are getting connected and communicating with each other are increasing at an exponential rate. The convenience of connecting to each other has come at a cost of sacrificing the security aspect. Due to that the number of Blackhat hacker are also increasing who gain access to a network illegally. Due to this the number of cyber attacks is also going up, with different types of techniques applied by them. Today having a firewall on the network is not enough, they cannot stop all the types of attacks coming from the external network. Intrusion Detection System plays an important role in obstructing these attacks at the entry of the network itself. This research paper talks about the new model of the classifier for the Intrusion Detection System. Two familiar classifiers, Random Forest and k-means clustering are used to develop the proposed model. The new technique increases the performance, accuracy and detection rate of the Intrusion Detection System. Every machine learning algorithm have their own advantages and disadvantages. NSL-KDD dataset has been used to train the proposed classifier model. Both Random Forest algorithm and K-means clustering algorithm are quite efficient in classifying the traffic data as normal or malicious when compared to others