Approximation and Error Prediction in Electrochemical Parameters Calculation Using Neural Networks

Abstract

Various interesting results have been achieved in calculation of electrochemical parameters in nanomaterials, using neural networks. There appear some error, during those calculations, and it varies depending on number of neurons in layers. In this research we deal with errors, calculated for neural networks with n=1,2…10, neurons in first or second layer. We applied mean square approximation method, in order to get explicite formula for predicton of error, for other cases.