STKAN: Kolmogorov-Arnold Networks for Spatio-Temporal Forecasting

2026-07-14

AI Summary

This work addresses the challenges of spatiotemporal forecasting in real-world traffic data, which arise from heterogeneous spatial correlations and nonlinear temporal dynamics. To tackle these issues, the authors propose a novel architecture integrating Kolmogorov–Arnold Networks (KANs) as powerful nonlinear function approximators. The model employs a learnable soft node grouping mechanism to capture spatial heterogeneity and extracts local temporal dependencies from compressed time series representations, complemented by a self-attention layer to model long-range temporal interactions. Extensive experiments on five established traffic forecasting benchmarks demonstrate that the proposed model significantly outperforms existing MLP-based variants, underscoring the critical role of KAN-based nonlinear modeling in enhancing spatiotemporal prediction performance.

Abstract

Real-world traffic data exhibit heterogeneous spatial correlations and nonlinear temporal dynamics, posing substantial challenges for accurate spatio-temporal forecasting. Existing approaches have developed increasingly sophisticated graph, attention, and decomposition architectures, while the influence of the underlying nonlinear function approximator has received comparatively less attention. In this work, we propose STKAN, a spatio-temporal forecasting architecture that introduces Taylor-polynomial Kolmogorov--Arnold Network modules into spatial and temporal token mixing. STKAN first constructs high-level spatial representations through a learnable soft node-group assignment mechanism, applies group-wise spatial mixing, and subsequently models temporal dependencies over the compressed sequence. Spatial and temporal self-attention layers are further employed to capture long-range interactions. Experiments on five traffic forecasting benchmarks show that STKAN achieves competitive performance and performs better than the evaluated MLP-based variant in the tested settings. These results suggest that the design of nonlinear function approximators can serve as a useful complement to architectural design in spatio-temporal forecasting.

Problem

• spatio-temporal forecasting
• heterogeneous spatial correlations
• nonlinear temporal dynamics
• traffic data

Innovation

• Kolmogorov-Arnold Networks
• spatio-temporal forecasting
• Taylor-polynomial KAN
• soft node-group assignment
• nonlinear function approximator