基于增量支持向量机算法的大数据网络安全系统检测技术  

Big data network security system detection technology based on incremental support vector machine algorithm

作　　者：王斌[1] 王业[1] 孙齐振 赵毅[1] WANG Bin;WANG Ye;SUN Qizhen;ZHAO Yi(Xinjiang Agricultural University,Urumqi 830052,China)

机构地区：[1]新疆农业大学,乌鲁木齐830052

出　　处：《自动化与仪器仪表》2024年第8期9-13,共5页Automation & Instrumentation

摘　　要：研究采用支持向量机算法来处理大规模网络流量数据的特性,使其能够适应网络流量数据特性、处理高维度以及分析非线性可分数据。研究提出并应用增量支持向量机算法解决大数据网络安全系统检测中的异常检测问题,算法综合利用了历史数据中的非支持向量样本和支持向量样本,以提高模型的性能和效率。研究结果显示,研究提出的算法在训练时间上明显优于传统算法,但分类准确率相近。最终,研究提出的算法在增量学习结束后的检测模型性能上优于传统算法,检测时间为传统算法的60.68%,异常检测率提升12.41%,误报率下降52.94%。因此,研究提出的安全系统检测技术在网络安全领域具有重要的应用前景。Study the characteristics of using support vector machine algorithm to process large-scale network traffic data,and ensure the classification accuracy of the model and the relationship between support vectors through association with incremental support vector machines.The study proposed and applied the incremental support vector machine algorithm to solve the anomaly detection problem in big data network security system detection.The algorithm comprehensively utilizes non-support vector samples and support vector samples in historical data to improve the performance and efficiency of the model.The research results show that the algorithm proposed in the study is significantly better than the traditional algorithm in terms of training time,but the classification accuracy is similar.In the end,the algorithm proposed in the study outperformed the traditional algorithm in terms of detection model performance after incremental learning.The detection time was 60.68%of the traditional algorithm,the anomaly detection rate increased by 12.41%,and the false alarm rate decreased by 52.94%.Therefore,the security system detection technology proposed in the study has important application prospects in the field of network security.

关 键 词：CSV-KKT-ISVM算法 网络安全检测 增量学习 网络异常流量 

分 类 号：TM77[电气工程—电力系统及自动化] TP39[自动化与计算机技术—计算机应用技术]