M3-Net: A Cost-Effective Graph-Free MLP-Based Model for Traffic Prediction

2025-08-11

AI Summary

Existing traffic forecasting models often rely on explicit graph structures or complex spatiotemporal attention mechanisms, resulting in high parameter counts and computational overhead, hindering practical deployment. To address this, we propose M3-Net—a lightweight, graph-free, pure MLP architecture. Methodologically, M3-Net innovatively integrates the Mixture of Experts (MoE) mechanism into the MLP-Mixer framework, augmented with temporal and spatiotemporal embeddings to enable efficient and scalable spatiotemporal modeling—entirely without graph neural networks. This design drastically reduces both parameter count and computational cost. Empirically, M3-Net achieves superior prediction accuracy with significantly fewer parameters across multiple real-world traffic datasets. It demonstrates strong generalization capability and low inference latency. Overall, M3-Net provides an efficient, end-to-end, and production-ready solution for large-scale intelligent transportation systems.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Computer Vision: Large Vision Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Achieving accurate traffic prediction is a fundamental but crucial task in the development of current intelligent transportation systems.Most of the mainstream methods that have made breakthroughs in traffic prediction rely on spatio-temporal graph neural networks, spatio-temporal attention mechanisms, etc. The main challenges of the existing deep learning approaches are that they either depend on a complete traffic network structure or require intricate model designs to capture complex spatio-temporal dependencies. These limitations pose significant challenges for the efficient deployment and operation of deep learning models on large-scale datasets. To address these challenges, we propose a cost-effective graph-free Multilayer Perceptron (MLP) based model M3-Net for traffic prediction. Our proposed model not only employs time series and spatio-temporal embeddings for efficient feature processing but also first introduces a novel MLP-Mixer architecture with a mixture of experts (MoE) mechanism. Extensive experiments conducted on multiple real datasets demonstrate the superiority of the proposed model in terms of prediction performance and lightweight deployment.

Problem

• Traffic prediction without relying on complete network structure
• Reducing complexity in capturing spatio-temporal dependencies
• Enabling efficient deployment on large-scale datasets

Innovation

• Graph-free MLP-based model for traffic prediction
• MLP-Mixer architecture with mixture of experts
• Time series and spatio-temporal embeddings