V.I. Kozik and E.S. Nejevenko (Russia)
dynamic processes, neural networks, adaptation, images sequence
The problems of prediction of dynamic processes propa gation at the Earth surface, described by the images se quence, are considered in this paper. The basic difference of the offered method from known is the opportunity of process modeling in the areas with sites, on which there are no conditions for process propagation. It’s reached by building the model combining the principles of the neural networks and macroneurons. This approach allows fulfill ing model adaptation using an actual process results.
Important Links:
Go Back
