An adaptive intrusion detection model based on ART2 neural network

作者：Liu Jifen^*

来源：Journal of Computational Information Systems, 2013, 9(19): 7775-7782.

DOI：10.12733/jcis6925

• Adaptive learning
• ART-2 neural networks
• Basic principles
• Detection accuracy
• Intrusion detection method
• Intrusion detection models
• Intrusion Detection Systems
• On-line detection

To improve the poor capability of on-line adaptive learning existing in the intrusion detection system and alert the detected intrusion attack according to its type so as to process later, an on-line adaptive intrusion detection model in light of the basic principle of ART2 neural network is proposed. It could remember the former learnt pattern stably, learn new pattern from coming connection records adaptively, recognize not only known types of attacks but also unknown types of attacks, and improve the effectiveness of intrusion detection effectively. Experiments demonstrated that this intrusion detection model effectively improved detection accuracy and decreased false alarm rate compared with the static learning intrusion detection method based on SVM.

• 出版日期2013
• 单位上海海事大学; 东华大学