A0334
Title: Privacy-preserving transfer learning for community detection using locally distributed multiple networks
Authors: Shujie Ma - University of California-Riverside (United States) [presenting]
Abstract: A new spectral clustering-based method is introduced, called TransNet, for transfer learning in community detection of network data. The goal is to improve the clustering performance of the target network using auxiliary source networks, which are locally stored across various sources, privacy-preserved, and heterogeneous. Notably, the source networks are allowed to have distinct privacy-preserving and heterogeneity levels that often happen in practice. To better utilize the information from the heterogeneous and privacy-preserved source networks, a new adaptive weighting method is proposed to aggregate the eigenspaces of the source networks and a regularization method that can automatically combine the weighted average eigenspace of the source networks with the eigenspace of the target network to achieve an optimal balance between them. It is also demonstrated that TransNet performs better than both the estimator only using the target network and the estimator using the weighted source networks.
