Scholarlink

<p>Intrusion Detection Systems are essential for networks to ensure the confidentiality, integrity, and availability of data. In today’s increasingly interconnected world, IDS remain a critical component of cybersecurity strategies. Traditional IDS, and even recent Deep Learning based approaches, often suffer from high false positive rates, computational inefficiencies, and limited adaptability. To address these challenges, We proposed a LAD-based IDS model in which we introduced a binarization technique that efficiently extracts critical patterns from heterogeneous and imbalanced network traffic data. The proposed model is evaluated using well-known datasets, such as NSL-KDD and KDD-Cup99, to ensure a rigorous assessment against diverse attack scenarios. Experimental evaluation reveals that the IDS-LAD-based method achieved an accuracy of 80.026% and 99.872%, precision of 80.921% and 99.870%, recall of 81.260% and 99.782%, and an F2-Score of 80.498% and 99.865% on the NSL-KDD and KDD-Cup99 datasets, respectively.</p>

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

IDS-LAD: Intrusion detection system using logical analysis of data

  • 期刊论文

  • 发表日期：

    2025年10月9日

查看全文
  • Bhaumikaditya Guleria,
  • Maroti Deshmukh,
  • Rakhi Nautiyal

摘要

Intrusion Detection Systems are essential for networks to ensure the confidentiality, integrity, and availability of data. In today’s increasingly interconnected world, IDS remain a critical component of cybersecurity strategies. Traditional IDS, and even recent Deep Learning based approaches, often suffer from high false positive rates, computational inefficiencies, and limited adaptability. To address these challenges, We proposed a LAD-based IDS model in which we introduced a binarization technique that efficiently extracts critical patterns from heterogeneous and imbalanced network traffic data. The proposed model is evaluated using well-known datasets, such as NSL-KDD and KDD-Cup99, to ensure a rigorous assessment against diverse attack scenarios. Experimental evaluation reveals that the IDS-LAD-based method achieved an accuracy of 80.026% and 99.872%, precision of 80.921% and 99.870%, recall of 81.260% and 99.782%, and an F2-Score of 80.498% and 99.865% on the NSL-KDD and KDD-Cup99 datasets, respectively.