HIDS-DT: An Effective Hybrid Intrusion Detection System Based on Decision Tree

Abstract

A hybrid intrusion detection approach combing both misuse detection and anomaly detection can detect newly discovered attacks while maintaining a relatively high detection rate. This paper presents a novel hybrid intrusion detection system based on protocol analysis and decision tree algorithms. Performance evaluation of the proposed system is conducted using Generalized Stochastic Petri Nets (GSPN). Simulation results show that this hybrid system can reach a high detection rate.