Compression of the geophysical data with application of vejvlet-transformation and neural networks

Abstract

In this paper the algorithm of a choice of parametres of vejvlet-transformation for compression of the geophysical data taking into account the set level of distortions is presented. The technique of construction of neural networks of the minimum complexity for compression of a vejvlet-spectrum of the geophysical data, providing quantity reduction synapse scales and acceleration of process of training of a network at the expense of decomposition for each class of the geophysical data of set of inputs and the first layer of neural network taking into account vejvlettransformation parametres is offered. The complex algorithm of compression of the geophysical data with application of vejvlet-transformation and neural networks is developed.