Privacy-robustness sensitivity analysis for secure development of differentially privacy deep learning under distribution shift
International Journal on Information Technologies and Security · 01.06.2026 · 18 (2) · 71-82
DOI: 10.59035/lbbt3945 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal on Information Technologies and Security, online erschienen am 01.06.2026. Verfasst von B. Lavanya und 1 weiteren Autor:innen.
Abstract (Originalsprache)
The widespread use of deep learning on sensitive data has positioned Differential Privacy (DP) as a central mechanism for protecting training confidentiality. Existing evaluations emphasize the privacy-utility trade-off measured on clean data, overlooking robustness under distribution shift. This study introduces Privacy-Robustness Sensitivity (PRS), a metric that explicitly quantifies the degradation in robustness induced by differential privacy by measuring the relative loss in corruption robustness with respect to clean accuracy as privacy constraints tighten. A privacy–utility–robustness evaluation framework is validated on CIFAR-10, CIFAR-10-C, CelebA, and a systematically constructed CelebA-C biometric corruption benchmark, enabling robustness-aware evaluation of privacy-preserving deep learning models for secure and resilient deployment scenarios.
Zitieren
APA
Lavanya, B., & Janani, S. (2026). Privacy-robustness sensitivity analysis for secure development of differentially privacy deep learning under distribution shift. International Journal on Information Technologies and Security, 18(2), 71-82. https://doi.org/10.59035/lbbt3945
Vancouver
Lavanya B, Janani S. Privacy-robustness sensitivity analysis for secure development of differentially privacy deep learning under distribution shift. International Journal on Information Technologies and Security. 2026;18(2):71-82. doi:10.59035/lbbt3945
BibTeX
@article{Lavanya2026,
author = {Lavanya, B. and Janani, S.},
title = {Privacy-robustness sensitivity analysis for secure development of differentially privacy deep learning under distribution shift},
journal = {International Journal on Information Technologies and Security},
year = {2026},
volume = {18},
number = {2},
pages = {71--82},
doi = {10.59035/lbbt3945}
}
EUVIMED