微型电脑应用2025，Vol.41Issue(4)：17-20,24,5.
基于GOOSE-SVM的网络安全智能预警方法研究
Research on Intelligent Early Warning Method of Network Security Based on GOOSE-SVM
摘要
Abstract
Aiming at the problem that the accuracy of the traditional network security intelligent early warning method is not high,a network security intelligent early warning method based on goose algorithm-support vector machine(GOOSE-SVM)is proposed.In this method,GOOSE optimization algorithm is used to optimize the penalty coefficient and kernel function param-eters of SVM to improve the prediction performance of SVM.The proposed method is applied to 4 types of network attack data in CICIDS2017 dataset,and compared with SVM and genetic algorithm-suppert vector machine(GA-SVM)early warning methods.The results show that the same early warning model has certain differences in different types of network attack detec-tion,and the proposed GOOSE-SVM network security intelligent early warning model has better early warning performance.This has certain practical guiding value for effectively improving the accuracy rate of intelligent early warning of network securi-ty and ensuring network security.关键词
网络安全/智能预警/鹅优化算法—支持向量机/核函数/惩罚系数Key words
network security/intelligent early warning/GOOSE-SVM/kernel function/penalty coefficient分类
信息技术与安全科学引用本文复制引用
何岩,王聪,李岳峰..基于GOOSE-SVM的网络安全智能预警方法研究[J].微型电脑应用,2025,41(4):17-20,24,5.基金项目
河北省自然科学基金项目(22HB295704) （22HB295704）
