Scholarlink

With the rapid expansion of network infrastructure, network intrusions are becoming more frequent, sophisticated, and volatile. Addressing the challenge of detecting intrusions in such extensive networks is both critical and complex. This proposed model introduces innovative deep learning methodologies to create a reliable intrusion detection mechanism aimed at identifying malicious attacks. The devised framework harnesses advanced deep learning methodologies, specifically integrating a fusion of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks, coupled with seven diverse optimizer functions. Performance evaluation is conducted utilizing the CIS IDS dataset, encompassing multi-attack classification tasks. The proposed model exhibits unparalleled performance compared to the Adamax optimizer, showcasing supremacy in both accuracy and loss metrics. Benchmarking against existing models, the LSTM-GRU model attains an exceptional accuracy rate of 99.09%. This signifies a notable enhancement of 1.07% in accuracy over preceding methodologies, reaffirming the efficacy and robustness of the proposed approach.

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

Cyber Vigilance Nexus: Advancing Intrusion Detection Network Accuracy Through Deep Learning Optimization

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Chetan Gupta,
  • Amit Kumar,
  • Neelesh Kumar Jain

摘要

With the rapid expansion of network infrastructure, network intrusions are becoming more frequent, sophisticated, and volatile. Addressing the challenge of detecting intrusions in such extensive networks is both critical and complex. This proposed model introduces innovative deep learning methodologies to create a reliable intrusion detection mechanism aimed at identifying malicious attacks. The devised framework harnesses advanced deep learning methodologies, specifically integrating a fusion of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks, coupled with seven diverse optimizer functions. Performance evaluation is conducted utilizing the CIS IDS dataset, encompassing multi-attack classification tasks. The proposed model exhibits unparalleled performance compared to the Adamax optimizer, showcasing supremacy in both accuracy and loss metrics. Benchmarking against existing models, the LSTM-GRU model attains an exceptional accuracy rate of 99.09%. This signifies a notable enhancement of 1.07% in accuracy over preceding methodologies, reaffirming the efficacy and robustness of the proposed approach.