Scholarlink

Privacy and the protection of the client’s data is at the foremost demand nowadays. The rapid rise in the variety of Android malware has also become a major concern for academicians and researchers. In order to protect client’s sensitive information, in this paper, a framework named as FLANdroid-A Privacy-Preserving framework to detect Malware for Android system using Semi-supervised algorithm based on Federated Learning that provide security to user data is proposed. To deal with the problems of non-IID data distribution and huge unlabelled dataset we used consistency regularisation and pseudo labeling to leverage semi-supervised learning mechanisms to improve model generalisation. Outcomes of the proposed framework are compared with dynamic analysis-based frameworks and accuracy of 95.3% and 0.93 F-measure is achieved.

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

FLANdroid—A Privacy-Preserving Framework to Detect Malware for Android System Using Semi-supervised Algorithm Based on Federated Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Pooja,
  • Arvind Mahindru,
  • Pardeep Kumar Arora

摘要

Privacy and the protection of the client’s data is at the foremost demand nowadays. The rapid rise in the variety of Android malware has also become a major concern for academicians and researchers. In order to protect client’s sensitive information, in this paper, a framework named as FLANdroid-A Privacy-Preserving framework to detect Malware for Android system using Semi-supervised algorithm based on Federated Learning that provide security to user data is proposed. To deal with the problems of non-IID data distribution and huge unlabelled dataset we used consistency regularisation and pseudo labeling to leverage semi-supervised learning mechanisms to improve model generalisation. Outcomes of the proposed framework are compared with dynamic analysis-based frameworks and accuracy of 95.3% and 0.93 F-measure is achieved.