Researchers develop self-supervised learning for Android malware detection

Researchers have developed a new method for detecting Android malware that addresses temporal bias in machine learning models. By constructing a time-stamped dataset and implementing a timestamp-verification procedure, their framework ensures models are evaluated based on actual app release times. The system utilizes self-supervised pre-training with BYOL to learn robust representations, achieving 98% accuracy and 89% F1 score under time-aware evaluation. The dataset and source code have been released to promote further research. AI

影响 Improves robustness of AI-based malware detection systems by addressing temporal bias.

排序理由 Academic paper introducing a new methodology and dataset for Android malware detection.

在 arXiv cs.LG 阅读 →

• Android
• MITRE ATT&CK
• VirusTotal

• 论文
• 安全

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →