Statistical analysis of CIDDS-001 dataset for Network Intrusion Detection Systems using Distance-based Machine Learning

A lot of research is being done on the development of effective Network Intrusion Detection Systems. Anomaly based Network Intrusion Detection Systems are preferred over Signature based Network Intrusion Detection Systems because of their better significance in detecting novel attacks. The research on the datasets being used for training and testing purpose in the detection model is equally concerned as better dataset quality can advance offline Intrusion Detection. Benchmark datasets like KDD99 and NSL-KDD cup 99 are outdated and face some major issues, which make them unsuitable for evaluating Anomaly based Network Intrusion Detection Systems. This paper presents the statistical analysis of labelled flow based CIDDS-001 dataset using k-nearest neighbour classification and k-means clustering algorithms. The analysis is done with respect to some prominent evaluation metrics used for evaluating Network Intrusion Detection Systems including Detection Rate, Accuracy and False Positive Rate.