VFL-RPS: Relevant Participant Selection in Vertical Federated Learning

TL;DR

The paper tackles the problem of selecting informative participants in vertical federated learning to reduce communication and computation without sacrificing performance. It introduces VFL-RPS, a privacy-preserving pre-training approach that securely computes feature relevance via Spearman correlations using secure multi-party computation and then applies forward selection to mitigate redundancy among active and passive parties. Through experiments on six datasets covering regression and classification, the method demonstrates that using about half of the participants can achieve performance comparable to using all, while outperforming existing VFL participant-selection baselines. The work provides a practical, privacy-conscious strategy to improve VFL efficiency and robustness in the presence of overlap or noise across party features.

Abstract

Federated Learning (FL) allows collaboration between different parties, while ensuring that the data across these parties is not shared. However, not every collaboration is helpful in terms of the resulting model performance. Therefore, it is an important challenge to select the correct participants in a collaboration. As it currently stands, most of the efforts in participant selection in the literature have focused on Horizontal Federated Learning (HFL), which assumes that all features are the same across all participants, disregarding the possibility of different features across participants which is captured in Vertical Federated Learning (VFL). To close this gap in the literature, we propose a novel method VFL-RPS for participant selection in VFL, as a pre-training step. We have tested our method on several data sets performing both regression and classification tasks, showing that our method leads to comparable results as using all data by only selecting a few participants. In addition, we show that our method outperforms existing methods for participant selection in VFL.

Figures (3)

• Figure 1: Vertical Federated Learning Setup
• Figure 2: Test accuracy and runtime with different number of passive parties
• Figure 3: VFL-RPS Workflow