摘要
将医院网络入侵行为作为研究对象,提出基于SSA和ELM的网络入侵特征选择模型,有效实施网络入侵行为检测。该方法应用SSA算法优选网络入侵特征属性,用于改进ELM网络分类性能,通过减少模型输入特征数,来降低计算复杂度。将模型用于医院网络Dos,Probe,R2L等攻击行为样本检测,结果表明检测准确率能够达90%以上,检测时长在0.5s以内,误报率不超0.3%,能满足医院网络入侵检测高效、准确、可靠的检测要求。
Taking hospital network intrusion behavior as the research object,a network intrusion Feature selection model based on SSA and ELM is proposed to effectively implement network intrusion behavior detection.This method applies the SSA algorithm to optimize network intrusion feature attributes,which is used to improve the classification performance of ELM networks.By reducing the number of input features in the model,the computational complexity is reduced.The model is applied to detect attack behavior samples such as Dos,Probe,and R2L in hospital networks.The results show that the detection accuracy can reach over 90%,the detection time is within 0.5 seconds,and the false alarm rate does not exceed o.3%.It can meet the requirements of efficient,accurate,and reliable detection in hospital network intrusion detection.
作者
杨威
YANG Wei(Suzhou First People's Hospital,Suzhou,Anhui 234000,China)
出处
《计算机应用文摘》
2023年第15期101-104,共4页
Chinese Journal of Computer Application
