Functional data privacy works better when smoothing comes first

On the Order of Privacy and Smoothing in Functional Mean Estimation

Cryptography and Security

Summary

When protecting sensitive data that changes over time, like health signals, it matters whether you smooth out the data patterns before or after adding privacy noise. The authors show that smoothing before privacy noise gives more accurate estimates when there is little measurement error. However, if there’s a lot of noise in the data, adding privacy noise first can work better. They also extend privacy guarantees to broader mathematical settings beyond typical spaces used before.

What this means in practice

  • For health data engineers: Improve algorithms that protect patient monitoring data by smoothing signals before applying privacy noise to reduce estimation errors.
  • For financial data analysts: Enhance privacy techniques for time series financial data streams using the recommended order of smoothing and privatization to optimize accuracy.

Authors

Haotian Chen, Yu-Wei Chen, Qifan Song, Jordan Awan

Abstract

Differential privacy (DP) is a leading framework for privacy protection, yet privacy mechanisms tailored to functional data analysis (FDA) remain relatively underexplored. For functional mean estimation, we investigate a fundamental question regarding the ordering of privatization and estimation by comparing two canonical pipelines: Fit-Then-Protect (FTP), which smooths trajectories before privatization, and Protect-Then-Fit (PTF), which privatizes observations at discrete time points before smoothing. When data are observed without error, we prove that FTP achieves strictly smaller mean squared error than PTF asymptotically (as the sample size increases) under the same $μ$-Gaussian differential privacy ($μ$-GDP) budget in both finite-dimensional Hilbert spaces and infinite-dimensional reproducing kernel Hilbert spaces (RKHS). Simulation studies support our theory and explore settings with observation error and fitting constraints, demonstrating that FTP outperforms PTF under small observation noise, whereas PTF can become preferable in high-noise settings. In addition to this comparison, we establish a novel GDP guarantee for Gaussian perturbations taking values in a Banach space, extending existing results beyond the RKHS setting.