摘要
传统的网络安全检测技术难以解决新网络流量无标签、未知攻击以及标签稀缺等问题。为此,将人工智能技术应用到上述场景的网络安全入侵检测中,使用无标签网络安全检测方案解决目标网络域和源网络域间特征分布不同的问题。针对未知网络攻击的问题构建一个未知攻击网络安全检测模型,针对标签稀缺问题将半监督学习和主动学技术结合起来,构建一种标签稀缺网络安全检测算法.
Traditional network security detection technologies are difficult to solve the problems of unlabeled new network traffic,unknown attacks and label scarcity.Therefore,artificial intelligence technology is applied to the network security intrusion detection of above scenarios,and unlabeled network security detection solutions are used to solve the problem of the different feature distributions between the target network domain and the source net⁃work domain.For the problem of unknown network attacks,a network security detection model for unknown at⁃tacks is constructed,and for the problem of label scarcity,semi-supervised learning and active learning techniques are combined to construct a network security detection algorithm for label scarcity.
作者
虢莉娟
GUO Lijuan(Yiyang Open University,Yiyang,Hunan Province,413000 China)
出处
《科技资讯》
2024年第11期21-23,共3页
Science & Technology Information
关键词
网络安全
人工智能
入侵检测
数据集
Network security
Artificial intelligence
Intrusion detection
Dataset