1. Science
  2. Publications
  3. Information Processing Systems
  4. 5(112)'2013
  5. All parameters of the evolving radial-basis neural function networks adaptive learning

All parameters of the evolving radial-basis neural function networks adaptive learning

E.V. Bodyanskiy, A.A. Deineko, M.Z. Stolnikova
Annotations languages:

In this work the architecture and method of learning artificial evolving radial-basis function neural networks that adjusts not only their weight, but also automatically determines the number of neurons, the location of centers of radial basis functions and parameters of the receptive field in the on-line mode with high speed and operation-data is proposed.
Keywords: artificial neuron network, adaptive teaching, radially-base neuron network, organize oneself map of T. Kokhonen, activating function of V. Epanechnikov