Scholarlink

As the world rushes into an era of an exponentially growing networked infrastructure and digital infrastructure, cyber security threats such as intrusion attempts, malware, and denial-of-service (DoS) have become more refined. Intrusion detection systems (IDSs) traditionally are not able to identify new and sophisticated threats in real-time though. This research presents a hybrid deep learning model, which combines the Convolutional Neural network with the Long Short Term Memory network for a better detection rate of network intrusion. The proposed CNN-LSTM architecture is tested on NSL-KDD dataset which exhibits superior performance when considering precision, recall and F1-score when compared to traditional machine learning classifiers.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The Problem of Improving Network Intrusion Detection Using a Hybrid CNN LSTM Deep Learning Model: A Cyber Security Perspective

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jatin Kumar,
  • Ashwani Kr. Verma

摘要

As the world rushes into an era of an exponentially growing networked infrastructure and digital infrastructure, cyber security threats such as intrusion attempts, malware, and denial-of-service (DoS) have become more refined. Intrusion detection systems (IDSs) traditionally are not able to identify new and sophisticated threats in real-time though. This research presents a hybrid deep learning model, which combines the Convolutional Neural network with the Long Short Term Memory network for a better detection rate of network intrusion. The proposed CNN-LSTM architecture is tested on NSL-KDD dataset which exhibits superior performance when considering precision, recall and F1-score when compared to traditional machine learning classifiers.