The 6th Spatial Data Science Symposium
Published November 20, 2025 | Version v1
Conference paper Open

ST-SplitVFL: Spatio-Temporal Split Vertical Federated Learning

Creators

  • 1. AIT Austrian Institute of Technology
  • 2. ROR icon Austrian Institute of Technology

Description

This paper introduces ST-SplitVFL, a novel framework for Spatio-Temporal Split Vertical Federated Learning, which extends vertical FL with split learning and spatial graph constraints. In this setting, clients maintain their local data while selectively exchanging encoded intermediate representations with geographically or logically connected peers. 

Files

SDSS25__ST_SplitVFL_2025-11-20.pdf

Files (888.6 kB)

Name Size Download all
md5:2eac4233272fcac9253fb8115c955b01
888.6 kB Preview Download

Additional details

Funding

European Commission
EMERALDS - Extreme-scale Urban Mobility Data Analytics as a Service 101093051