Scholarlink

Zero-day attacks, which take advantage of unidentified software flaws to compromise patient data and interrupt vital medical services, are becoming a big menace to the healthcare industry. These attacks frequently result in ransomware outbreaks, medical record manipulation, unauthorised access to sensitive data, and disruptions to regular healthcare operations. Given that it depends so heavily on medical devices and networked digital systems, the healthcare industry is especially vulnerable. To address this challenge, this study proposes a machine learning-based solution by integrating an advanced Stacking Classifier with a K-Means Clustering algorithm for real-time detection and prediction of zero-day attacks, achieving an accuracy of 97.86% on completely unknown attacks and 80.04% on mixed datasets. This approach enhances cybersecurity in healthcare, protecting patient data and ensuring the continuity of critical medical operations.

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

Machine Learning-Based Hybrid Solution for Zero-Day Attack Prediction in Healthcare: A Stacking and Clustering Approach

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Prajwal V Shenoy,
  • M Prerana,
  • R Sanath Kumar,
  • Dhyey Patel,
  • V Sarasvathi

摘要

Zero-day attacks, which take advantage of unidentified software flaws to compromise patient data and interrupt vital medical services, are becoming a big menace to the healthcare industry. These attacks frequently result in ransomware outbreaks, medical record manipulation, unauthorised access to sensitive data, and disruptions to regular healthcare operations. Given that it depends so heavily on medical devices and networked digital systems, the healthcare industry is especially vulnerable. To address this challenge, this study proposes a machine learning-based solution by integrating an advanced Stacking Classifier with a K-Means Clustering algorithm for real-time detection and prediction of zero-day attacks, achieving an accuracy of 97.86% on completely unknown attacks and 80.04% on mixed datasets. This approach enhances cybersecurity in healthcare, protecting patient data and ensuring the continuity of critical medical operations.