An automatic differentiation system for the age of differential privacy

Usynin, Dmitrii; Ziller, Alexander; Knolle, Moritz; Trask, Andrew; Prakash, Kritika; Rueckert, Daniel; Kaissis, Georgios

doi:10.48550/arxiv.2109.10573

Cited by 2 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, its correct application is complicated by the introduction of unintuitive parameters like ε or δ [19,53], or by the requirement to understand additional DP definitions like node, edge or graph-level DP. Thus, besides systems which automate sensitivity calculations and the application of DP to generic machine learning workflows [113], works similar to [18] are required, which investigate user expectations and interpretations of DP, paving the way for an improved user experience for practitioners.…”

Section: Challenges and Outlookmentioning

confidence: 99%

Differentially Private Guarantees for Analytics and Machine Learning on Graphs: A Survey of Results

Mueller,

Usynin,

Paetzold

et al. 2024

JPC

View full text Add to dashboard Cite

We study the applications of differential privacy (DP) in the context of graph-structured data and discuss the formulations of DP applicable to the publication of graphs and their associated statistics as well as machine learning on graph-based data, including graph neural networks (GNNs). Interpreting DP guarantees in the context of graph-structured data can be challenging, as individual data points are interconnected (often non-linearly or sparsely). This connectivity complicates the computation of individual privacy loss in differentially private learning. The problem is exacerbated by an absence of a single, well-established formulation of DP in graph settings. This issue extends to the domain of GNNs, rendering private machine learning on graph-structured data a challenging task. A lack of prior systematisation work motivated us to study graph-based learning from a privacy perspective. In this work, we systematise different formulations of DP on graphs, discuss challenges and promising applications, including the GNN domain. We compare and separate works into graph analytics tasks and graph learning tasks with GNNs. We conclude our work with a discussion of open questions and potential directions for further research in this area.

show abstract

Section: Challenges and Outlookmentioning

confidence: 99%

Differentially Private Guarantees for Analytics and Machine Learning on Graphs: A Survey of Results

Mueller,

Usynin,

Paetzold

et al. 2024

JPC

View full text Add to dashboard Cite

show abstract

“…Moreover, its correct application is complicated by the introduction of unintuitive parameters like ε or δ [105,106], or by the requirement to understand additional DP definitions like node, edge or graph-level DP. Thus, besides systems which automate sensitivity calculations and the application of DP to generic machine learning workflows [107], works similar to [108] are required, which investigate user expectations and interpretations of DP, paving the way for an improved user experience for practitioners.…”

Section: Interpretability Of Dp In Graphsmentioning

confidence: 99%

SoK: Differential Privacy on Graph-Structured Data

Mueller¹,

Usynin²,

Paetzold³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In this work, we study the applications of differential privacy (DP) in the context of graph-structured data. We discuss the formulations of DP applicable to the publication of graphs and their associated statistics as well as machine learning on graph-based data, including graph neural networks (GNNs). The formulation of DP in the context of graph-structured data is difficult, as individual data points are interconnected (often non-linearly or sparsely). This connectivity complicates the computation of individual privacy loss in differentially private learning. The problem is exacerbated by an absence of a single, well-established formulation of DP in graph settings. This issue extends to the domain of GNNs, rendering private machine learning on graph-structured data a challenging task. A lack of prior systematisation work motivated us to study graph-based learning from a privacy perspective. In this work, we systematise different formulations of DP on graphs, discuss challenges and promising applications, including the GNN domain. We compare and separate works into graph analysis tasks and graph learning tasks with GNNs. Finally, we conclude our work with a discussion of open questions and potential directions for further research in this area.

show abstract

An automatic differentiation system for the age of differential privacy

Cited by 2 publications

References 12 publications

Differentially Private Guarantees for Analytics and Machine Learning on Graphs: A Survey of Results

Differentially Private Guarantees for Analytics and Machine Learning on Graphs: A Survey of Results

SoK: Differential Privacy on Graph-Structured Data

Contact Info

Product

Resources

About