基于孤立森林算法的计算机网络信息安全漏洞检测方法
Computer Network Information Security Vulnerability Detection Method Based on Isolated Forest Algorithm
刘悦1
作者信息
- 1. 中电科航空电子有限公司,四川成都 611731
- 折叠
摘要
为解决现有的计算机网络安全检测方法存在的误报率较高问题,文章提出基于孤立森林算法的计算机网络信息安全漏洞检测方法.对计算机网络信息安全数据集进行预处理,剔除样本噪声清洗数据,对数据特征与类别间的相对熵进行计算;确定网络攻击行为,构建子森林潜在攻击检测块,设定阈值判断进入滑动窗口的数据是否为异常数据,实现计算机网络信息安全漏洞检测.实验结果表明,实验组的召回精度均在95％以上,F1均在0.68以上.这表明该方法能够对漏洞进行精准识别,提升漏洞的分类精度,漏洞检测效果较好.
Abstract
In order to solve the problem of high false alarm rate of existing computer network security detection methods,this paper proposes computer network information security vulnerability detection method based on isolated forest algorithm.The data set of computer network information security is preprocessed,the sample noise is removed to clean the data,and the relative entropy between data features and categories is calculated.The network attack behavior is determined,the potential attack detection block of sub-forest is constructed,the threshold is set to determine whether the data entering the sliding window is abnormal data,and the computer network information security vulnerability detection is realized.The experimental results show that the recall accuracy of the experimental group is above 95％,and the Fl is above 0.68.This shows that this method can accurately identify vulnerabilities,improve the classification accuracy of vulnerabilities,and achieve good vulnerability detection results.
关键词
孤立森林算法/计算机网络/信息安全/漏洞检测Key words
isolated forest algorithm/computer network/information security/vulnerability detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据