计算机应用 ›› 2005, Vol. 25 ›› Issue (01): 144-146.DOI: 10.3724/SP.J.1087.2005.0144

• 信息安全 • 上一篇    下一篇

AdaBoost方法在入侵检测技术上的应用

郭红刚1,2,方敏1,2   

  1. 1.综合业务网国家重点实验室;2.西安电子科技大学计算机学院
  • 发布日期:2005-01-01 出版日期:2005-01-01
  • 基金资助:

    综合业务网国家重点实验室开放基金资助项目(ISN6 7)

Application of AdaBoost method in IDS

GUO Hong-gang1, 2, FANG Min1, 2   

  • Online:2005-01-01 Published:2005-01-01

摘要: 详细介绍了AdaBoost方法的原理与算法,以及如何将AdaBoost方法应用于入侵检测的核心技术———入侵分析技术中,并且用实验数据对该方法进行验证,同时将AdaBoost的测试结果与基分类器BP的测试结果进行比较,得出AdaBoost方法能够大幅度提高弱分类器识别率,以及它适合于IDS数据的结论。

关键词: 入侵检测, Boosting, AdaBoost, 分类器, 样本, 训练, 测试

Abstract: Firstly, the principle and algorithm of AdaBoost method were described in detail. Secondly, the application of this method to intrusion analytical technique that was the core of the technique of intrusion detection was introduced. Finally, how to verify this method with experimental data was shown. Meanwhile, compared with that of base classifiers BP, the test result of AdaBoost method could remarkably increase the accuracy of weak classer and was fit for the IDS data set.

Key words:  intrusion detection, Boosting, AdaBoost, classer, sample, train, test

中图分类号: