Modern computational intelligence and data science applications frequently operate on complex, non-Euclidean domains. Applying differential privacy (DP) to these structures requires mechanisms that respect their underlying geometry. Instance-adaptive mechanisms in DP improve utility over global worst-case bounds but often require heavy-tailed noise, leading to numerical instability and unbounded variance. We introduced a geometric framework that lifted sensitivity analysis to the tangent bundle of a Riemannian manifold. The proposed directional local sensitivity operator captured maximal directional derivatives along minimizing geodesics, enabling our operator normed adaptive Laplace mechanism (ON–ALM). By enforcing a $ \beta $-log-Lipschitz stability condition on the regularized sensitivity, noise was precisely calibrated via directional operator norms. Under $ (\varepsilon, \delta) $-DP, extracting a typical set allowed the use of light-tailed Laplace noise, achieving strictly bounded variance, improved numerical stability, and high utility for geometric and directional queries.
Citation: Bekir Danış. Operator normed adaptive Laplace mechanism: A manifold-based approach for differential privacy[J]. Electronic Research Archive, 2026, 34(9): 6043-6071. doi: 10.3934/era.2026267
Modern computational intelligence and data science applications frequently operate on complex, non-Euclidean domains. Applying differential privacy (DP) to these structures requires mechanisms that respect their underlying geometry. Instance-adaptive mechanisms in DP improve utility over global worst-case bounds but often require heavy-tailed noise, leading to numerical instability and unbounded variance. We introduced a geometric framework that lifted sensitivity analysis to the tangent bundle of a Riemannian manifold. The proposed directional local sensitivity operator captured maximal directional derivatives along minimizing geodesics, enabling our operator normed adaptive Laplace mechanism (ON–ALM). By enforcing a $ \beta $-log-Lipschitz stability condition on the regularized sensitivity, noise was precisely calibrated via directional operator norms. Under $ (\varepsilon, \delta) $-DP, extracting a typical set allowed the use of light-tailed Laplace noise, achieving strictly bounded variance, improved numerical stability, and high utility for geometric and directional queries.
| [1] | C. Dwork, F. McSherry, K. Nissim, A. Smith, Calibrating noise to sensitivity in private data analysis, in Theory of Cryptography Conference (TCC), Springer, Berlin, Heidelberg, (2006), 265–284. |
| [2] | K. Nissim, S. Raskhodnikova, A. Smith, Smooth sensitivity and sampling in private data analysis, in Proceedings of the 39th Annual ACM Symposium on Theory of Computing, (2007), 75–84. |
| [3] | D. Durfee, Unifying Laplace mechanism with instance optimality in differential privacy, preprint, arXiv: 2505.02798. |
| [4] |
J. Dong, A. Roth, W. J. Su, Gaussian differential privacy, J. R. Stat. Soc. B, 84 (2022), 3–37. https://doi.org/10.1111/rssb.12454 doi: 10.1111/rssb.12454
|
| [5] | M. Reimherr, J. Awan, KNG: The K-norm gradient mechanism, in Advances in Neural Information Processing Systems, 32 (2019). |
| [6] |
A. Han, B. Mishra, P. Jawanpuria, J. Gao, Differentially private Riemannian optimization, Mach. Learn., 113 (2024), 1133–1161. https://doi.org/10.1007/s10994-023-06508-5 doi: 10.1007/s10994-023-06508-5
|
| [7] |
C. Li, S. Long, H. Liu, Y. Choi, H. Sekiya, Z. Li, Enhancing sparse mobile crowdsensing with manifold optimization and differential privacy, IEEE Trans. Inf. Forensics Secur., 19 (2024), 6070–6083. https://doi.org/10.1109/TIFS.2024.3407668 doi: 10.1109/TIFS.2024.3407668
|
| [8] | Z. Huang, W. Huang, P. Jawanpuria, B. Mishra, Federated learning on Riemannian manifolds with differential privacy, preprint, arXiv: 2404.10029. |
| [9] |
S. Garg, V. Torra, Privacy in manifolds: Combining k-anonymity with differential privacy on Fréchet means, Comput. Secur., 144 (2024), 103983. https://doi.org/10.1016/j.cose.2024.103983 doi: 10.1016/j.cose.2024.103983
|
| [10] | P. He, L. Tang, M. A. Rahimian, J. Joshi, Conformal-DP: Differential privacy on Riemannian manifolds via conformal transformation, preprint, arXiv: 2504.20941. |
| [11] | M. P. do Carmo, Riemannian Geometry, Birkhäuser, Boston, 1992. |
| [12] | J. M. Lee, Introduction to Riemannian Manifolds, 2nd edition, Springer, Cham, 2018. |
| [13] | G. Casella, R. L. Berger, Statistical Inference, 2nd edition, Thomson Learning, Belmont, CA, 2002. |
| [14] |
S. Dasgupta, A. Gupta, An elementary proof of a theorem of Johnson and Lindenstrauss, Random Struct. Algorithms, 22 (2003), 60–65. https://doi.org/10.1002/rsa.10073 doi: 10.1002/rsa.10073
|
| [15] | C. Dwork, G. N. Rothblum, S. Vadhan, Boosting and differential privacy, in Proceedings of the 51st Annual IEEE Symposium on Foundations of Computer Science, (2010), 51–60. |
Figures(1) / Tables(4)
Bekir Danış. Operator normed adaptive Laplace mechanism: A manifold-based approach for differential privacy[J]. Electronic Research Archive, 2026, 34(9): 6043-6071. doi: 10.3934/era.2026267
DownLoad: