Complete Undirected Graph Augmented Bayes Classifier and its Application in Intrusion Detection System
Shifu Chen · 2008
The strong independence assumption made by the nave Bayes classifier supposes that every attribute is independent from the rest of the attributes given the state of the class variable.This assumption rarely holds true in the intrusion detection datasets.This paper models a new algorithm based on the complete undirected Graph Augmented Bayes classifier (GAB) that takes into account all influences of attributes to reduce the nave Bayes independence assumption.We conduct experiments by using MIT intrusion detection datasets.The experimental results show that the new algorithm results in a significant improvement in detection accuracy.