Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

HSTFL: A Heterogeneous Federated Learning Framework for Misaligned Spatiotemporal Forecasting

Shuowei Cai, Hao Liu

TL;DR

This paper proposes a Heterogeneous SpatioTemporal Federated Learning (HSTFL) framework to enable multiple clients to collaboratively harness geo-distributed time series data from different domains while preserving privacy, and proposes a cross-client virtual node alignment block to incorporate cross-client spatiotemporal dependencies via a multi-level knowledge fusion scheme.

Abstract

Spatiotemporal forecasting has emerged as an indispensable building block of diverse smart city applications, such as intelligent transportation and smart energy management. Recent advancements have uncovered that the performance of spatiotemporal forecasting can be significantly improved by integrating knowledge in geo-distributed time series data from different domains, \eg enhancing real-estate appraisal with human mobility data; joint taxi and bike demand predictions. While effective, existing approaches assume a centralized data collection and exploitation environment, overlooking the privacy and commercial interest concerns associated with data owned by different parties. In this paper, we investigate multi-party collaborative spatiotemporal forecasting without direct access to multi-source private data. However, this task is challenging due to 1) cross-domain feature heterogeneity and 2) cross-client geographical heterogeneity, where standard horizontal or vertical federated learning is inapplicable. To this end, we propose a Heterogeneous SpatioTemporal Federated Learning (HSTFL) framework to enable multiple clients to collaboratively harness geo-distributed time series data from different domains while preserving privacy. Specifically, we first devise vertical federated spatiotemporal representation learning to locally preserve spatiotemporal dependencies among individual participants and generate effective representations for heterogeneous data. Then we propose a cross-client virtual node alignment block to incorporate cross-client spatiotemporal dependencies via a multi-level knowledge fusion scheme. Extensive privacy analysis and experimental evaluations demonstrate that HSTFL not only effectively resists inference attacks but also provides a significant improvement against various baselines.

HSTFL: A Heterogeneous Federated Learning Framework for Misaligned Spatiotemporal Forecasting

TL;DR

This paper proposes a Heterogeneous SpatioTemporal Federated Learning (HSTFL) framework to enable multiple clients to collaboratively harness geo-distributed time series data from different domains while preserving privacy, and proposes a cross-client virtual node alignment block to incorporate cross-client spatiotemporal dependencies via a multi-level knowledge fusion scheme.

Abstract

Spatiotemporal forecasting has emerged as an indispensable building block of diverse smart city applications, such as intelligent transportation and smart energy management. Recent advancements have uncovered that the performance of spatiotemporal forecasting can be significantly improved by integrating knowledge in geo-distributed time series data from different domains, \eg enhancing real-estate appraisal with human mobility data; joint taxi and bike demand predictions. While effective, existing approaches assume a centralized data collection and exploitation environment, overlooking the privacy and commercial interest concerns associated with data owned by different parties. In this paper, we investigate multi-party collaborative spatiotemporal forecasting without direct access to multi-source private data. However, this task is challenging due to 1) cross-domain feature heterogeneity and 2) cross-client geographical heterogeneity, where standard horizontal or vertical federated learning is inapplicable. To this end, we propose a Heterogeneous SpatioTemporal Federated Learning (HSTFL) framework to enable multiple clients to collaboratively harness geo-distributed time series data from different domains while preserving privacy. Specifically, we first devise vertical federated spatiotemporal representation learning to locally preserve spatiotemporal dependencies among individual participants and generate effective representations for heterogeneous data. Then we propose a cross-client virtual node alignment block to incorporate cross-client spatiotemporal dependencies via a multi-level knowledge fusion scheme. Extensive privacy analysis and experimental evaluations demonstrate that HSTFL not only effectively resists inference attacks but also provides a significant improvement against various baselines.
Paper Structure (53 sections, 4 theorems, 21 equations, 9 figures, 3 tables)

This paper contains 53 sections, 4 theorems, 21 equations, 9 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Federated Learning on Spatiotemporal Data
  4. Federated Learning on Heterogeneous Data
  5. Multi-source Data for Urban Spatiotemporal Forecasting
  6. Preliminary
  7. Problem Formulation
  8. Threat Model
  9. Heterogeneous Spatiotemporal Federated Learning
  10. Overview
  11. Vertical Federated Spatiotemporal Representation Learning
  12. Temporal Representation Learning
  13. Vertical Federated Spatial Representation Learning
  14. Cross-Client Virtual Node Alignment
  15. Virtual Node Generation
  16. ...and 38 more sections

Key Result

theorem 1

The Gaussian mechanism defined in Definition def:gaussian preserves $(\epsilon ,\delta)$-DP for each publication step dp.

Figures (9)

  • Figure 1: Illustrative example of multi-source private spatiotemporal data hold by multiple clients. Each client holds a distinct type of private geo-distributed time series (i.e., spatiotemporal data). These geo-distributed time series data in each clients are also misaligned in geographical location.
  • Figure 2: Overview of the HSTFL framework.
  • Figure 3: The virtual node alignment module.
  • Figure 4: HSTFL with different local spatiotemporal module.
  • Figure 5: The model performance and attack result of the White-box attack with differential privacy.
  • ...and 4 more figures

Theorems & Definitions (7)

  • Definition 1: The Gaussian Mechanism
  • theorem 1
  • Definition 2: inverse function
  • Definition 3: information leakage
  • theorem 2
  • Lemma 1
  • theorem 2