Scholarlink

This research aims to improve cyber threat detection in IoT systems using advanced machine learning techniques. Existing methods struggle with the complex, high-dimensional nature of IoT traffic data. The proposed approach integrates Extreme Gradient Boosting (XGBoost), optimized with novel parameter tuning, to effectively identify and mitigate IoT-based cyber threats. Using the University of New Brunswick CIC IoT dataset, the model achieved 93.19% accuracy, outperforming traditional methods like KNN, SVM, and Random Forest. These results highlight the model's robustness and potential for enhancing real-time IoT threat detection, contributing to a more secure IoT ecosystem.

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

Detecting and Mitigating IoT-Based Cyber Threats Through Machine Learning on Network Traffic Data

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Raju Bomma,
  • Kiran L. N. Eranki

摘要

This research aims to improve cyber threat detection in IoT systems using advanced machine learning techniques. Existing methods struggle with the complex, high-dimensional nature of IoT traffic data. The proposed approach integrates Extreme Gradient Boosting (XGBoost), optimized with novel parameter tuning, to effectively identify and mitigate IoT-based cyber threats. Using the University of New Brunswick CIC IoT dataset, the model achieved 93.19% accuracy, outperforming traditional methods like KNN, SVM, and Random Forest. These results highlight the model's robustness and potential for enhancing real-time IoT threat detection, contributing to a more secure IoT ecosystem.