UniMamba: A Unified Spatial-Temporal Modeling Framework with State-Space and Attention Integration

2026-03-06

AI Summary

This work addresses the challenges of modeling complex temporal dependencies and cross-variable interactions in multivariate time series forecasting by proposing a unified spatiotemporal framework that integrates state space models with attention mechanisms. The method uniquely combines the Mamba state space model and attention in a deeply integrated architecture, incorporating variable-specific channel encoding, spatiotemporal attention, and a feedforward temporal dynamics layer. It further enhances global dependency capture by integrating FFT-Laplace transforms with Temporal Convolutional Networks (TCNs). Extensive experiments on eight public benchmark datasets demonstrate that the proposed approach significantly outperforms current state-of-the-art methods in both prediction accuracy and computational efficiency.

Abstract

Multivariate time series forecasting is fundamental to numerous domains such as energy, finance, and environmental monitoring, where complex temporal dependencies and cross-variable interactions pose enduring challenges. Existing Transformer-based methods capture temporal correlations through attention mechanisms but suffer from quadratic computational cost, while state-space models like Mamba achieve efficient long-context modeling yet lack explicit temporal pattern recognition. Therefore we introduce UniMamba, a unified spatial-temporal forecasting framework that integrates efficient state-space dynamics with attention-based dependency learning. UniMamba employs a Mamba Variate-Channel Encoding Layer enhanced with FFT-Laplace Transform and TCN to capture global temporal dependencies, and a Spatial Temporal Attention Layer to jointly model inter-variate correlations and temporal evolution. A Feedforward Temporal Dynamics Layer further fuses continuous and discrete contexts for accurate forecasting. Comprehensive experiments on eight public benchmark datasets demonstrate that UniMamba consistently outperforms state-of-the-art forecasting models in both forecasting accuracy and computational efficiency, establishing a scalable and robust solution for long-sequence multivariate time-series prediction.

Problem

• multivariate time series forecasting
• temporal dependencies
• cross-variable interactions
• long-sequence prediction
• computational efficiency

Innovation

• state-space model
• attention mechanism
• multivariate time series forecasting
• spatial-temporal modeling
• Mamba