基于信息增益的贝叶斯态势要素提取

A Method Using Information Gain and Naive Bayes to Extract Network Situation Information

针对网络安全态势要素提取、约简、分类问题,文章对收集的网络数据,通过计算每个属性的信息增益约简,并通过信息增益设置权值,获得关联性强的态势因子。在朴素贝叶斯进行的安全态势因子分类中,对朴素贝叶斯算法进行了改进,加入了调控因子,提高了分类效果,并实现了对恶意攻击的检测。最后,用入侵检测数据集对改进的方法进行测试,并将得到的结果与传统的方法进行比较,得到了更好的效果。

Network security situational factors of extraction exist some problems m reduction and classification. We collect network data to obtain the strong correlation between situational factors through setting up right value. Through the improvement of naive bayesian analysis,the detection of malicious attack was realized. This paper tries to experiment with the intrusion detection data set, and compares the traditional method to the better effect.