Comparison of Anomaly Detection Accuracy of Host-based Intrusion Detection Systems based on Different Machine Learning Algorithms
Yukyung Shin, Kangseok Kim
Abstract
Among the different host-based intrusion detection systems, an anomaly-based intrusion detection system detects attacks based on deviations from normal behavior; however, such a system has a low detection rate. Therefore, several studies have been conducted to increase the accurate detection rate of anomaly-based intrusion detection systems; recently, some of these studies involved the development of intrusion detection models using machine learning algorithms to overcome the limitations of existing anomaly-based intrusion detection methodologies as well as signature-based intrusion detection methodologies. In a similar vein, in this study, we propose a method for improving the intrusion detection accuracy of anomaly-based intrusion detection systems by applying various machine learning algorithms for classification of normal and attack data. To verify the effectiveness of the proposed intrusion detection models, we use the ADFA Linux Dataset which consists of system call traces for attacks on the latest operating systems. Further, for verification, we develop models and perform simulations for host-based intrusion detection systems based on machine learning algorithms to detect and classify anomalies using the Arena simulation tool.