Estudi dels algoritmes d'aprenentatge per a xarxes neurals amb valors sinàptics discrets

    Project Details

    Description

    The objectives of this project are the development of a neural network emulation board that will be able to be connected to a PC; the development of a software package to make easier the use of the emulation board and the development of a demonstrator to prove the real applicability of the system. The board emulation process presents high performances for their use with feedback neural networks. In this case, our algorithm reduces the computing complexity to O(N). Theoretical studies and simulations on conventional computers have been already done. To complete the process, a learning algorithm for discrete weights will be developed. As previously, studies and simulations have been already done. Two researchers teams will work in the project: the first belonging to the Computer Science Dpt. of the UAB, and the second one belonging to the CNM. The current importance of the project is justi
    StatusFinished
    Effective start/end date6/08/916/08/94

    Funding

    • Comisión Interministerial de Ciencia y Tecnología (CICYT): €26,684.94

    Fingerprint Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.