基于SVM-DT-MLP模型的Web日志异常流量检测研究
Research on Web Log Abnormal Traffic Detection Based on the SVM-DT-MLP Model
魏璐露 1程楠楠1
作者信息
- 1. 江西科技学院 信息工程学院,江西 南昌 330029
- 折叠
摘要
随着Web应用程序的普及,网络攻击和安全漏洞的风险日益增加.Web日志文件详细记录了网站运行信息,对日志中的流量进行分类从而检测出异常攻击流量是保障网页长期提供稳定、安全服务行之有效的方法之一.文中将Voting特征选择与Stacking集成相结合,构建了SVM-DT-MLP模型,并将其用于Web日志异常流量检测.测试结果表明,SVM-DT-MLP模型的性能显著优于单一算法模型,其Precision(精确度)达到 92.44％,Recall(召回率)达到 92.43％,F1-Score(F1 值)达到 92.44％.这意味着该模型能够有效地检测出异常攻击流量,并在保障网页提供稳定和安全服务方面具有很好的效果.
Abstract
With the popularity of Web applications,the risk of cyber attacks and security vulnerabilities is increasing.Web log files record the running information of websites in detail.Classifying the traffic in logs to detect abnormal attack traffic is one of the effective methods to ensure the long-term stability and security service provided by Web pages.In this paper,Voting feature selection and the Stacking integration are combined to construct the SVM-DT-MLP model,and it is used to detect abnormal traffic in Web logs.The test results show that the performance of SVM-DT-MLP model is significantly better than that of the single algorithm model,with the precision reaching 92.44％,the recall reaching 92.43％and the F1-Score reaching 92.44％.This means that the model can effectively detect abnormal attack traffic and has a good effect in ensuring stable and secure services provided by Web pages.
关键词
Web日志/异常流量检测/Stacking集成/Voting特征选择/机器学习Key words
Web log/abnormal traffic detection/Stacking integration/Voting feature selection/Machine Learning引用本文复制引用
基金项目
江西省教育厅科学技术研究项目(GJJ2202609)
出版年
2024
国家科技期刊平台
NSTL
万方数据