SpikeSTAG: Spatial-Temporal Forecasting via GNN-SNN Collaboration

2025/08/04 by Baoxin Hu, Hu, Bang, Changze Lv +13
Computer Science · Engineering · #Advanced Graph Neural Networks #Advanced Memory and Neural Computing #Artificial Intelligence (cs.AI) #Artificial neural network #Block (permutation group theory) #FOS: Computer and information sciences #Feature (linguistics) #Feature extraction #Graph #Machine Learning (cs.LG) #Sequence (biology) #Spike (software development) #Spiking neural network #Traffic Prediction and Management Techniques

paper · pdf · doi:10.48550/arxiv.2508.02069

published in arXiv (Cornell University) (Cornell University)

openalex publication_date 2025/08/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Spiking neural networks (SNNs), inspired by the spiking behavior of biological neurons, offer a distinctive approach for capturing the complexities of temporal data. However, their potential for spatial modeling in multivariate time-series forecasting remains largely unexplored. To bridge this gap, we introduce a brand new SNN architecture, which is among the first to seamlessly integrate graph structural learning with spike-based temporal processing for multivariate time-series forecasting. Specifically, we first embed time features and an adaptive matrix, eliminating the need for predefined graph structures. We then further learn sequence features through the Observation (OBS) Block. Building upon this, our Multi-Scale Spike Aggregation (MSSA) hierarchically aggregates neighborhood information through spiking SAGE layers, enabling multi-hop feature extraction while eliminating the need for floating-point operations. Finally, we propose a Dual-Path Spike Fusion (DSF) Block to integrate spatial graph features and temporal dynamics via a spike-gated mechanism, combining LSTM-processed sequences with spiking self-attention outputs, effectively improve the model accuracy of long sequence datasets. Experiments show that our model surpasses the state-of-the-art SNN-based iSpikformer on all datasets and outperforms traditional temporal models at long horizons, thereby establishing a new paradigm for efficient spatial-temporal modeling.

Citations

• TS-LIF: A Temporal Segment Spiking Neuron Network for Time Series Forecasting
• STAA-SNN: Spatial-Temporal Attention Aggregator for Spiking Neural Networks
• SpikingSSMs: Learning Long Sequences with Sparse and Parallel Spiking State Space Models
• SpikeGraphormer: A High-Performance Graph Transformer with Spiking Graph Attention
• FourierGNN: Rethinking Multivariate Time Series Forecasting from a Pure Graph Perspective
• Unleashing the Potential of Spiking Neural Networks for Sequential Modeling with Contextual Embedding
• Scaling Up Dynamic Graph Representation Learning via Spiking Neural Networks
• Exploiting Spiking Dynamics with Spatial-temporal Feature Normalization in Graph Learning
• Optimal ANN-SNN Conversion for Fast and Accurate Inference in Deep Spiking Neural Networks
• Deep Residual Learning in Spiking Neural Networks
• Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting
• Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting
• Multivariate Time Series Classification Using Spiking Neural Networks
• Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks
• Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks
• A CNN–LSTM model for gold price time-series forecasting
• Graph WaveNet for Deep Spatial-Temporal Graph Modeling
• AliGraph: A Comprehensive Graph Neural Network Platform
• Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition
• Inductive Representation Learning on Large Graphs
• Structural-RNN: Deep Learning on Spatio-Temporal Graphs
• Networks of spiking neurons: The third generation of neural network models

Related

• Efficient and Effective Time-Series Forecasting with Spiking Neural Networks
• TS-LIF: A Temporal Segment Spiking Neuron Network for Time Series Forecasting
• STAA-SNN: Spatial-Temporal Attention Aggregator for Spiking Neural Networks
• Spiking Neural Networks for Temporal Processing: Status Quo and Future Prospects
• A lightweight Spatial-Temporal Graph Neural Network for Long-term Time Series Forecasting