Abstract: Bayes classifier model is a powerful tool for classifying attack types in intrusion detection. This paper proposes an improved Bayesian algorithm model, which improves the traditional Bayesian algorithm and weakens the strong independence relation assumption of Naive Bayes. Then, it offers an experimental study and analysis, which shows that this improved Bayes classifier algorithm model enhances the classified precision of the intrusion detection. Finally, it points out the direction of the further study.

Key words: Bayes classifier; Intrusion detection; Classified accuracy

摘要： 贝叶斯分类模型是入侵检测中用于攻击类型分类的有力工具。在总结前人成果的基础上，提出了一个改进的贝叶斯模型，对朴素贝叶斯算法进行了改进，降低了朴素贝叶斯算法的强独立性假设，提高了入侵检测的分类精度，并通过试验对算法进行了验证和性能分析。同时，指出了下一步的研究方向。

关键词: 贝叶斯分类器；入侵检测；分类精度