• DocumentCode
    3630024
  • Title

    Linking and Activation Potential Optimization in the Pulse Coupled Neural Network

  • Author

    R. Forgac;I. Mokris

  • Author_Institution
    Institute of Informatics, Slovak Academy of Sciences, Bratislava, Slovakia, radoslav.forgac@savba.sk
  • fYear
    2008
  • Firstpage
    85
  • Lastpage
    88
  • Abstract
    The paper introduces an approach for linking and activation potential optimization in the Pulse Coupled Neural Network (PCNN) to reduce the number of PCNN parameters. Linking potential has high influence on control of internal activity of PCNN neuron. The neuron internal activity, represented by activation potential, determines the pulse response mode of neuron in the PCNN. Because standard PCNN has up to ten parameters, it is very difficult set up all optimal values of them. With linking and activation potential optimization are better possibilities for PCNN application because four parameters were eliminated.
  • Keywords
    "Joining processes","Neural networks","Neurons","Mathematical model","Image recognition","Biological system modeling","Brain modeling","Pixel","Pulse generation","Image analysis"
  • Publisher
    ieee
  • Conference_Titel
    Computational Cybernetics, 2008. ICCC 2008. IEEE International Conference on
  • Print_ISBN
    978-1-4244-2874-8
  • Type

    conf

  • DOI
    10.1109/ICCCYB.2008.4721384
  • Filename
    4721384