Differentially Private Multivariate Time Series Forecasting of Aggregated Human Mobility With Deep Learning: Input or Gradient Perturbation?

TL;DR

This paper compares input and gradient perturbation methods for differentially private deep learning models in multivariate human mobility forecasting, showing minimal performance loss and providing a privacy-preserving solution for urban planning applications.

Contribution

It introduces a comprehensive comparison of input and gradient perturbation techniques in deep learning for privacy-preserving mobility forecasting, with extensive experiments on real data.

Findings

Differentially private models perform nearly as well as non-private models.

Performance loss ranges from 0.57% to 2.8%.

Gradient and input perturbation methods yield similar results.

Abstract

This paper investigates the problem of forecasting multivariate aggregated human mobility while preserving the privacy of the individuals concerned. Differential privacy, a state-of-the-art formal notion, has been used as the privacy guarantee in two different and independent steps when training deep learning models. On one hand, we considered \textit{gradient perturbation}, which uses the differentially private stochastic gradient descent algorithm to guarantee the privacy of each time series sample in the learning stage. On the other hand, we considered \textit{input perturbation}, which adds differential privacy guarantees in each sample of the series before applying any learning. We compared four state-of-the-art recurrent neural networks: Long Short-Term Memory, Gated Recurrent Unit, and their Bidirectional architectures, i.e., Bidirectional-LSTM and Bidirectional-GRU. Extensive experiments were conducted with a real-world multivariate mobility dataset, which we published openly along with this paper. As shown in the results, differentially private deep learning models trained under gradient or input perturbation achieve nearly the same performance as non-private deep learning models, with loss in performance varying between $0.57\%$ to $2.8\%$. The contribution of this paper is significant for those involved in urban planning and decision-making, providing a solution to the human mobility multivariate forecast problem through differentially private deep learning models.