Intrusion Detection System using Naive Bayes algorithm

The growth of internet usage increased the need of security in network which is monitored by Intrusion Detection System (IDS). Using machine learning algorithms is common for implementing any IDS to detect network traffic weather it is normal or attack. Naive Bayes algorithm is one of the popular supervised classification algorithm for categorical dataset which is built on conditional independence of feature assumption. Our experimental research focused on comparison of traditional Naive Bayes algorithm and PCA based implementation using with(sklearn) and without built in python library. Experimental results using PCA based NSL-KDD intrusion detection system indicate better accuracy compared to traditional Naive Bayes in both with and without built in sklearn python library.