Description: Intrusion detection is an important aspect of network security. Many current detection systems have trouble providing real-time network intrusion detection, particularly where very high dimensional data is involved. The intrusion detection system of the present invention provides real-time network intrusion detection by projecting the high dimensional dataset to a lower dimensional space using the random projection technique, then performing intrusion detection in the lower dimensional space using a support vector machine (SVM) classifier.
Detecting intrusions in the projected lower dimension reduces the complexity of the underlying algorithms, which makes it more suitable for real time detection. Moreover, lower dimensional data can be stored and transmitted more efficiently than its higher dimensional data, thereby saving system resources.
For more information please contact Geoffrey Pinski at 513-558-5696 or email@example.com