A privacy-by-design pipeline using static Drebin features fed to an SVM with dual-reject thresholds defers 6.7% of samples to sandboxed dynamic analysis and reaches F1 0.87 on 2024-2025 data without accessing user information.
In: 2020 IEEE 45th Conference on Local Computer Networks (LCN)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Don't Trust Us: A privacy-by-design android malware detection pipeline
A privacy-by-design pipeline using static Drebin features fed to an SVM with dual-reject thresholds defers 6.7% of samples to sandboxed dynamic analysis and reaches F1 0.87 on 2024-2025 data without accessing user information.