This page links to additional material for our ICML 2025 spotlight paper

Privacy Amplification by Structured Subsampling for Deep Differentially Private Time Series Forecasting
by Jan Schuchardt, Mina Dalirrooyfard, Jed Guzelkabaagac, Anderson Schneider, Yuriy Nevmyvaka, Stephan Günnemann
Published at International Conference on Machine Learning (ICML) 2025

Links

[Paper | Poster (t.b.d.) | Github ]

Abstract

Many forms of sensitive data, such as web traffic, mobility data, or hospital occupancy, are inherently sequential. The standard method for training machine learning models while ensuring privacy for units of sensitive information, such as individual hospital visits, is differentially private stochastic gradient descent (DP-SGD). However, we observe in this work that the formal guarantees of DP-SGD are incompatible with timeseries-specific tasks like forecasting, since they rely on the privacy amplification attained by training on small, unstructured batches sampled from an unstructured dataset. In contrast, batches for forecasting are generated by (1) sampling sequentially structured time series from a dataset, (2) sampling contiguous subsequences from these series, and (3) partitioning them into context and ground-truth forecast windows. We theoretically analyze the privacy amplification attained by this structured subsampling to enable the training of forecasting models with sound and tight event- and user-level privacy guarantees. Towards more private models, we additionally prove how data augmentation amplifies privacy in self-supervised training of sequence models. Our empirical evaluation demonstrates that amplification by structured subsampling enables the training of forecasting models with strong formal privacy guarantees.

Cite

Please cite our paper if you use the method in your own work:

@InProceedings{schuchardt2025_forecasting,
author = {Schuchardt, Jan and Dalirrooyfard, Mina and Guzelkabaagac, Jed and Schneider, Anderson and Nevmyvaka, Yuriy and G{\"u}nnemann, Stephan},
title = {Privacy Amplification by Structured Subsampling for Deep Differentially Private Time Series Forecasting},
booktitle = {International Conference on Machine Learning (ICML)},
year = {2025}
}