Node-Level Differentially Private Graph Neural Networks
Daigavane, Ameya, Madan, Gagan, Sinha, Aditya, Thakurta, Abhradeep Guha, Aggarwal, Gaurav, Jain, Prateek
–arXiv.org Artificial Intelligence
Graph Neural Networks (GNNs) are a popular technique for modelling graph-structured data and computing node-level representations via aggregation of information from the neighborhood of each node. However, this aggregation implies an increased risk of revealing sensitive information, as a node can participate in the inference for multiple nodes. This implies that standard privacy-preserving machine learning techniques, such as differentially private stochastic gradient descent (DP-SGD) - which are designed for situations where each data point participates in the inference for one point only - either do not apply, or lead to inaccurate models. In this work, we formally define the problem of learning GNN parameters with node-level privacy, and provide an algorithmic solution with a strong differential privacy guarantee. We employ a careful sensitivity analysis and provide a non-trivial extension of the privacy-by-amplification technique to the GNN setting. An empirical evaluation on standard benchmark datasets demonstrates that our method is indeed able to learn accurate privacy-preserving GNNs which outperform both private and non-private methods that completely ignore graph information.
arXiv.org Artificial Intelligence
Aug-26-2022
- Country:
- North America > United States
- New York > New York County
- New York City (0.04)
- California > Santa Clara County
- Palo Alto (0.04)
- Santa Clara (0.04)
- New York > New York County
- Asia > Myanmar
- Tanintharyi Region > Dawei (0.04)
- North America > United States
- Genre:
- Research Report (0.83)
- Industry:
- Information Technology > Security & Privacy (1.00)
- Technology: