基于深度学习的网络入侵检测技术研究  

Research on Network Intrusion Detection Technology Based on Deep Learning

作　　者：孟禹[1] MEMG Yu(School of Artificial Intelligence,Anshan Normal University,Anshan Liaoning 114007,China)

机构地区：[1]鞍山师范学院人工智能学院,辽宁鞍山114007

出　　处：《鞍山师范学院学报》2025年第2期56-60,共5页Journal of Anshan Normal University

摘　　要：随着深度学习技术的发展,基于深度学习的入侵检测技术已经成为网络安全领域的重要研究方向之一.提出一种结合卷积神经网络和长短期记忆网络的入侵检测模型,即CNN-LSTM模型.针对该模型对少数类样本识别率低的问题,利用K-Means聚类算法和Tomek Links方法对数据集进行数据平衡,设计出基于KMS-Tomek-CNN-LSTM的网络入侵检测模型,并在实验数据集上取得了较高的精确度和F 1值,验证了该模型的有效性.With the continuous progress of deep learning technology,intrusion detection technology based on deep learning has become an important research direction in the field of network security.In this paper,we propose an intrusion detection model that combines convolutional neural network and long short-term memory network,CNN-LSTM model.Also for the problem of low recognition rate of this model for a few classes of samples,data balancing of the datasets using K-Means and Tomek Links methods is used to design a network intrusion detection model based on KMS-Tomek-CNN-LSTM,and higher accuracy and F 1 Score are achieved on the experimental datasets,which verifies the validity of the method in this paper.

关 键 词：深度学习 入侵检测 卷积神经网络 长短期记忆网络 

分 类 号：TP393.2[自动化与计算机技术—计算机应用技术]