Abstract:Spatio-temporal mobility modeling plays a pivotal role in the advancement of mobile computing. Nowadays, data is frequently held by various distributed silos, which are isolated from each other and confront limitations on data sharing. Given this, there have been some attempts to introduce federated learning into spatio-temporal mobility modeling. Meanwhile, the distributional heterogeneity inherent in the spatio-temporal data also puts forward requirements for model personalization. However, the existing methods tackle personalization in a model-centric manner and fail to explore the data characteristics in various data silos, thus ignoring the fact that the fundamental cause of insufficient personalization in the model is the heterogeneous distribution of data. In this paper, we propose a novel distribution-oriented personalized Fed erated learning framework for Cro ss-silo S patio- T emporal mobility modeling (named FedCroST ), that leverages learnable spatio-temporal prompts to implicitly represent the local data distribution patterns of data silos and guide the local models to learn the personalized information. Specifically, we focus on the potential characteristics within temporal distribution and devise a conditional diffusion module to generate temporal prompts that serve as guidance for the evolution of the time series. Simultaneously, we emphasize the structure distribution inherent in node neighborhoods and propose adaptive spatial structure partition to construct the spatial prompts, augmenting the spatial information representation. Furthermore, we introduce a denoising autoencoder to effectively harness the learned multi-view spatio-temporal features and obtain personalized representations adapted to local tasks. Our proposal highlights the significance of latent spatio-temporal data distributions in enabling personalized federated spatio-temporal learning, providing new insights into modeling spatio-temporal mobility in data silo scenarios. Extensive experiments conducted on real-world datasets demonstrate that FedCroST outperforms the advanced baselines by a large margin in diverse cross-silo spatio-temporal mobility modeling tasks.

Modeling Spatio-Temporal Mobility Across Data Silos Via Personalized Federated Learning

Federated Representation Learning with Data Heterogeneity for Human Mobility Prediction

Hierarchical Federated Learning: the Interplay of User Mobility and Data Heterogeneity

FedPSE: Personalized Sparsification with Element-wise Aggregation for Federated Learning

FediOS: Decoupling Orthogonal Subspaces for Personalization in Feature-skew Federated Learning

Survey of Federated Learning Models for Spatial-Temporal Mobility Applications

Improving Global Generalization and Local Personalization for Federated Learning

Spatio-temporal Heterogeneous Federated Learning for Time Series Classification with Multi-view Orthogonal Training

An Adaptive Federated Relevance Framework for Spatial Temporal Graph Learning

Achieving Personalized Federated Learning with Sparse Local Models

Sparse Federated Learning with Hierarchical Personalized Models

HSTFL: A Heterogeneous Federated Learning Framework for Misaligned Spatiotemporal Forecasting

PMF: A Privacy-preserving Human Mobility Prediction Framework via Federated Learning

When Transfer Learning Meets Cross-City Urban Flow Prediction: Spatio-Temporal Adaptation Matters

FedCross: Towards Accurate Federated Learning via Multi-Model Cross-Aggregation

An Optimal Transport Approach to Personalized Federated Learning

Predicting Collective Human Mobility via Countering Spatiotemporal Heterogeneity

Factor-Assisted Federated Learning for Personalized Optimization with Heterogeneous Data

Personalized Federated Learning on Flowing Data Heterogeneity under Restricted Storage

FedPer++: Toward Improved Personalized Federated Learning on Heterogeneous and Imbalanced Data

Federated Learning for Mobility Applications