arXiv:2604.23025v2 Announce Type: replace-cross Abstract: Android malware detectors built with machine learning often suffer from temporal bias: models are trained and evaluated without respecting apps' actual release times, inflating accuracy and weakening real-world robustness. We address this by constructing a time-stamped dataset of benign and malicious Android apps and introducing a timestamp-verification procedure to ensure temporal accuracy. We then propose a detection framework that uses Bootstrap Your Own Latent (BYOL) for self-supervised pre-training to learn obfuscation-resilient re
The proliferation of Android devices and the increasing sophistication of malware necessitate more robust and adaptive detection methods, which traditional ML struggles to provide due to temporal bias.
This research addresses a critical vulnerability in cybersecurity by proposing a more resilient and temporally accurate method for detecting Android malware, crucial for safeguarding mobile ecosystems and user data.
Machine learning models for malware detection can become significantly more accurate and robust in real-world scenarios by integrating self-supervised learning and timestamp-verified datasets, mitigating the 'temporal bias' problem.
- · Android users
- · Cybersecurity firms
- · Financial services
- · Mobile app developers
- · Malware developers
- · Cyber attackers
- · Obsolete malware detection systems
Reduced rates of successful Android malware attacks and data breaches.
Increased trust in Android platforms and mobile banking/e-commerce applications.
Potential for similar self-supervised and time-stamped approaches to be adopted for other cybersecurity domains, raising the bar for attackers across various platforms.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG