Scholarlink

To address the growing demand for robust Network Intrusion Detection Systems (NIDS), we propose a stage-wise curriculum learning framework combined with image-based transformation techniques and Explainable AI (XAI). This approach leverages XAI for transparency and comprehensibility, focusing on adaptive learning and scalable detection. By training the model on increasingly complex attack scenarios, the proposed architecture facilitates efficient and precise optimization. The explanatory tool SHAP (SHapley Additive exPlanations) is integrated provide detailed insights into model predictions. The ensemble stacking and model-to-model interoperability further enhances detection efficiency and reliability. Experiments validate the effectiveness of the proposed method, achieving the following accuracies: 97% on the CIC-Apt-IIoT dataset, 96% on the Edge-IIoT dataset, and 92% on the CIC-IoV-2024 dataset. This highlights the efficiency of the framework and setting a new benchmark for NIDS.

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

Curriculum Learning with Image Transformation and Explainable AI for Improved Network Intrusion Detection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sathwik Narkedimilli,
  • C. Pavan Kumar,
  • Raghavendra Ramachandra

摘要

To address the growing demand for robust Network Intrusion Detection Systems (NIDS), we propose a stage-wise curriculum learning framework combined with image-based transformation techniques and Explainable AI (XAI). This approach leverages XAI for transparency and comprehensibility, focusing on adaptive learning and scalable detection. By training the model on increasingly complex attack scenarios, the proposed architecture facilitates efficient and precise optimization. The explanatory tool SHAP (SHapley Additive exPlanations) is integrated provide detailed insights into model predictions. The ensemble stacking and model-to-model interoperability further enhances detection efficiency and reliability. Experiments validate the effectiveness of the proposed method, achieving the following accuracies: 97% on the CIC-Apt-IIoT dataset, 96% on the Edge-IIoT dataset, and 92% on the CIC-IoV-2024 dataset. This highlights the efficiency of the framework and setting a new benchmark for NIDS.