The backpropagation neural network for modelling

Trần Ngọc Hà, Nguyễn Thanh Thủy


In this paper we shall investigate some aspects on using the back propagation neural network for modeling. As an illustration, an experimental application to modelling equilibrium data of the liquid-liquid system Dy0HCl0DEHPA has been taken. It is shown that the neural network approach is an effective method for modelling various systems of this kind. Experimental results showed that errors of prediction of the proposed neural network arc really in the experimental range.


Journal of Computer Science and Cybernetics ISSN: 1813-9663

Published by Vietnam Academy of Science and Technology