Research
Privacy-Preserving Transfer Learning for Community Detection using Locally Distributed Multiple Networks
arXiv:2504.00890v2 Announce Type: replace-cross Abstract: Modern applications increasingly involve highly sensitive network data, where raw edges cannot be shared due to privacy constraints. We propos
arXiv:2504.00890v2 Announce Type: replace-cross Abstract: Modern applications increasingly involve highly sensitive network data, where raw edges cannot be shared due to privacy constraints. We propose exttt{TransNet}, a new spectral clustering-based transfer learning framework that improves community detection on a target network by leveraging heterogeneous, locally stored, and privacy-preserved auxiliary source networks. Our focus is the extit{local differential privacy} regime, in which each local data provider perturbs edges via extit{randomized response} before release, requiring no trusted third party. exttt{TransNet} aggregates source eigenspaces through a novel adaptive weighting scheme that accounts for both privacy and heterogeneity, and then regularizes the weighted source eigenspace with the target eigenspace to optimally balance the two. Theoretically, we establish an error-bound-oracle property: the estimation error for the aggregated eigenspace depends only on extit{informative sources}, ensuring robustness when some sources are highly heterogeneous or heavily privatized. We further show that the error bound of exttt{TransNet} is no greater than that of estimators using only the target network or only (weighted) sources. Empirically, exttt{TransNet} delivers strong gains across a range of privacy levels and heterogeneity patterns. For completeness, we also present exttt{TransNetX}, an extension based on Gaussian perturbation of projection matrices under the assumption that trusted local data curators are available.
Related
- Differentially Private and Federated Structure Learning in Bayesian Networks
- Poisoning with A Pill: Circumventing Detection in Federated Learning
- Decentralized Learning via Random Walk with Jumps
- SIGMA: An Efficient Heterophilous Graph Neural Network with Fast Global Aggregation
- Discrete Bayesian Sample Inference for Graph Generation
Source: arXiv cs.LG | 2026-04-15