A Hybrid Intrusion Detection System Based on C5.0 Decision Tree and One-Class SVM

Provided by: Inpressco
Topic: Data Management
Format: PDF
Cyber security threats have become increasingly sophisticated and complex. Intrusion detection which is one of the main problems in computer security has the main goal to detect infrequent access or attacks and to protect internal networks. A new hybrid intrusion detection method combining multiple classifiers for classifying anomalous and normal activities in the computer network is presented. The misuse detection model is built based on the C5.0 decision tree algorithm and using the information collected anomaly detection model is built which is implemented by one-class Support Vector Machine (SVM).

Find By Topic