Neural network modelling of cover thickness of contact surface by the complex of geophysical fields.

№1 (2018)

Nikitin A.A., Cheremisina E.N., Malinina C.C.

Abstract
It’s considered neural network modelling from the point of the sum of convolution of potential and electromagnetic original fields with weight factors that are identified on areas where the position of contact surface is found based on results of seismology. Final result of neural net formed from plenty of solutions of forward problems is submitted by minimization of generalized distance.

Section: Modeling geo objects and geo-processes

Keywords: neural net, sum of convolution, forward problems, generalized distance.