Comparative analysis of K-Means method and Naïve Bayes method for brute force attack visualization

This paper presents 2-Dimensional visualization to categorize packets of network traffic into normal data pattern and attack data pattern based on the patterns resulted by a brute force attack. Two clustering methods: K-Means and Naïve Bayes methods are used to produce the data to be visualized. Experiments using ISCX and DARPA dataset were conducted. Brute force assaults on some service protocols. This paper focuses on SSH service for ISCX dataset and TELNET service for DARPA dataset. Visual analysis of the experimental results show a better results in term of accuracy by reducing false alarms.