Most deep neural networks today optimize directly for performance on a task using only example input/output pairs. However, this excludes potential sources of knowledge which could improve performance. We propose a generalized frame-work for neural architectures which contain many semantically-meaningful variables. We call it the Multi-Semantic-Stage Neural Network (MSSNN).


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Multi -Semantic-Stage Neural Networks


    Beteiligte:


    Erscheinungsdatum :

    15.07.2024


    Format / Umfang :

    1227389 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Semantic and Logic Theories in Neural Networks

    Sugiyama, S. | British Library Conference Proceedings | 1994



    Semantic Communication System Based on Convolutional Neural Networks

    Wang, Jiawei / Jia, Xiaohui / Deng, Keyan | IEEE | 2023


    River Stage Forecasting Using Artificial Neural Networks

    Thirumalaiah, K. / Deo, M. C. | British Library Online Contents | 1998


    Early Stage Surge Discovery Based On Neural Networks

    Han, Z. / Qu, L. / Chen, Y. | British Library Online Contents | 1996