The author proposes a new Host-Based Hybrid Intrusion Detection System. The Intrusion Detection Analyzer Module consists of two analyzers that work in a hybrid architecture: Anomaly Detection Analyzer and Misuse Detection Analyzer. This way, the Anomaly Detection Analyzer is trained with attack-free session data and normal behaviour is learnt so it raises an alarm when it detects a deviation from this normal behaviour. Self Organizing Map, an unsupervised machine learning algorithm, is used. The Misuse Detection Analyzer uses a C4.5 Decision Tree. Finally, Decision Making Module decides whether the session is normal or an attack. The proposed hybrid architecture works very accurately. It is an essential book. Any professional can benefit from such a lecture.
I believe it will be very useful, interesting and inspiring for anyone who is performing research or is just interested in IDS, network security, information security, machine learning, neural networks, pattern recognition, and data mining. This book provides a lot of innovative ideas for students who are preparing thesis and dissertations in Information Security or Machine Learning domain. It would also be great for Computer Science and Electrical Engineering Departments to have this book in their library. ( )