基于Boosting集成学习算法的网络入侵检测技术研究  

Research on Network Intrusion Detection Technology Based on Boosting Integration Learning Algorithm

作　　者：易琳 王欣[1] Yi Lin;Wang Xin(School of Computer Science,Civil Aviation Flight University of China,Guanghan 618307)

机构地区：[1]中国民用航空飞行学院计算机学院,广汉618307

出　　处：《现代计算机》2021年第29期41-45,共5页Modern Computer

摘　　要：传统机器学习模型在网络入侵检测方面存在识别率低等问题,为了进一步提升检测率,主要应用Boosting集成学习算法进行相关的检测和研究,同时通过随机森林法来针对关键特征予以提取,构建多类分类器模型,同时借助KDD99数据来针对试验予以验证。实验结果表明Boosting集成学习算法能够较好的识别攻击类型数据。改进的GDBT,Xgboost相比随机森林效果较好,整体的准确率和召回率相比较优,表现出较高的预测精度。Research on intrusion detection related issues,today network attack model upgrade fast,more secretive encryption of traffic,in the past,traditional machine learning model for intrusion detection in low recognition rate problems,in order to further detec⁃tion rate for ascension,this article mainly used Boosting integration learning algorithm to expand testing analysis,At the same time,the random forest method was used to extract the key features,and the multi-class classifier model was constructed.Meanwhile,the KDD99 data was used to verify the experiment.The experimental results show that Boosting ensemble learning algorithm can identify attack type data well.Improved GDBT,XGBoost is better than Random Forest.Compared with the overall accuracy and recall rate,it shows a higher prediction accuracy.

关 键 词：网络入侵检测技术 Xgboost算法 梯度提升决策树 随机森林 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术] TP181[自动化与计算机技术—计算机科学与技术]