Open access
2026
HADAR-UAV: Risk-Calibrated One-Class Learning Framework for Zero-Day Intrusion Detection in Unmanned Aerial Vehicle Networks
This novel intrusion detection framework integrates masked autoencoder representation learning with Deep Support Vector Data Description under conformal prediction guarantees to calibrate risk and shows statistically significant improvement over current one-class methods.
C. Şahín, Siti Fatimah Abdul Razak, Arif Ullah et al.
· Computers, Materials & C... · 0 citations