Traffic forecasting is an important part of many future intelligent transportation systems and can be particularly useful for planning and navigation applications. The task is challenging because of the complex spatial patterns of road networks and dynamic temporal nature of traffic conditions. Recently, deep learning architectures with specially designed graph convolution layers to extract spatial patterns and recurrent or temporal convolution layers to extract temporal patterns have achieved good results for this task. In this paper, we propose a Mixture of Experts (MoE) based model integration framework to enhance the performance of these state-of-the-art traffic prediction models. In addition, we propose a novel entropy based loss function to improve the training of the MoE ensemble. Our experiments show that the performance of the Spatio-Temporal Graph Convolution Network (STGCN), a state of the art model, can be significantly improved.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Mixture of Experts based Model Integration for Traffic State Prediction


    Beteiligte:


    Erscheinungsdatum :

    01.06.2022


    Format / Umfang :

    745850 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Interpretable Cascading Mixture-of-Experts for Urban Traffic Congestion Prediction

    Jiang, Wenzhao / Han, Jindong / Liu, Hao et al. | ArXiv | 2024

    Freier Zugriff

    A Time Series is Worth Five Experts: Heterogeneous Mixture of Experts for Traffic Flow Prediction

    Wang, Guangyu / Chen, Yujie / Gao, Ming et al. | ArXiv | 2024

    Freier Zugriff

    Traffic speed forecasting by mixture of experts

    Coric, V. / Zhuang Wang, / Vucetic, S. | IEEE | 2011


    Traffic state prediction integration method

    QI YONG / YAN HE | Europäisches Patentamt | 2023

    Freier Zugriff