期刊名称：Oriental Journal of Computer Science and Technology
印刷版ISSN：0974-6471
出版年度：2010
卷号：3
期号：1
页码：165-170
出版社：Oriental Scientific Publishing Company
摘要：Network Intrusion Detection aims at distinguishing the behavior of the network. It is aninseparable part of the information security system. Due to rapid development of attack pattern it isnecessar y to develop a system which can upgrade itself as new threats are detected. Also detectionrate should be high because the rate with which attack is carried out on the network is very high. Inresponse to this problem AdaBoost Based Algorithm is proposed which has high detection rate as wellas low false alarm rate. In this algorithm decision stumps are used as weak classifier. The decisionrules are provided for both categorical and continuous features. Weak classifier for continuous featuresand weak classifier for categorical features are combined to form a strong classifier. Strategies foravoiding the over fitting are adopted to improve the performance of the algorithm.
