Intrusion Detection System (IDS) has increasingly become a crucial issue for computer and network systems. Optimizing performance of IDS becomes an important open problem which receives more and more attention from the research community. In this work, A multi-layer intrusion detection model is designed and developed to achieve high efficiency and improve the detection and classification rate accuracy .we effectively apply Machine learning techniques (C5 decision tree, Multilayer Perceptron neural network and Na\"ive Bayes) using gain ratio for selecting the best features for each layer as to use smaller storage space and get higher Intrusion detection performance. Our experimental results showed that the proposed multi-layer model using C5 decision tree achieves higher classification rate accuracy, using feature selection by Gain Ratio, and less false alarm rate than MLP and na\"ive Bayes. Using Gain Ratio enhances the accuracy of U2R and R2L for the three machine learning techniques (C5, MLP and Na\"ive Bayes) significantly. MLP has high classification rate when using the whole 41 features in Dos and Probe layers.
翻译:入侵探测系统(IDS)日益成为计算机和网络系统的一个关键问题。优化IDS的性能已成为一个重要的开放问题,引起了研究界越来越多的关注。在这项工作中,设计并开发了一个多层入侵探测模型,以实现高效,提高检测和分类率的准确性。我们有效地应用机器学习技术(C5决定树、多层过敏神经网络和Na\'ve Bayes),在选择每一层的最佳特征时使用较小的存储空间并获得更高的入侵探测性能方面采用增益率。我们的实验结果表明,使用C5决定树的拟议多层模型,使用增益率的特征选择,实现了更高的分类率准确性,比MLP和na\'Bayes低错误的警报率。对于三种机器学习技术(C5、MLP和Na\'ive Bayes),利用增益率提高U2R和R2L的准确性。在使用多斯和Probe层整个41个特征时,MLP的分类率很高。