HiMoE: Heterogeneity-Informed Mixture-of-Experts for Fair Spatial-Temporal Forecasting

TL;DR

This paper introduces HiMoE, a novel framework for fair spatial-temporal forecasting that addresses heterogeneity among nodes, achieving state-of-the-art accuracy and fairness through specialized graph convolution and mixture-of-experts modules.

Contribution

The paper proposes HiMoE, a heterogeneity-informed mixture-of-experts framework with a new graph convolution and benchmark for fairness in spatial-temporal forecasting.

Findings

Achieves at least 9.22% improvement over baselines.

Addresses trend and cardinality heterogeneity effectively.

Introduces STFairBench benchmark for fairness evaluation.

Abstract

Achieving both accurate and consistent predictive performance across spatial nodes is crucial for ensuring the validity and reliability of outcomes in fair spatial-temporal forecasting tasks. However, existing training methods treat heterogeneous nodes with a fully averaged perspective, resulting in inherently biased prediction targets. Balancing accuracy and consistency is particularly challenging due to the multi-objective nature of spatial-temporal forecasting. To address this issue, we propose a novel Heterogeneity-Informed Mixture-of-Experts (HiMoE) framework that delivers both uniform and precise spatial-temporal predictions. From a model architecture perspective, we design the Heterogeneity-Informed Graph Convolutional Network (HiGCN) to address trend heterogeneity, and we introduce the Node-wise Mixture-of-Experts (NMoE) module to handle cardinality heterogeneity across nodes. From an evaluation perspective, we propose STFairBench, a benchmark that handles fairness in spatial-temporal prediction from both training and evaluation stages. Extensive experiments on four real-world datasets demonstrate that HiMoE achieves state-of-the-art performance, outperforming the best baseline by at least 9.22% across all evaluation metrics.