ARTIFICIAL NEURAL NETWORK METHOD APPLIED ON THE NONLINEAR MULTIVARIATE PROBLEMS

This paper presents the research of the possibility to apply Artificial Neural Network (ANN) for solving nonlinear multivariate problems of the technological processes. After concurrent analysis facilitated using multivariate regression analysis (MRA) and ANN, on the same set of data, aiming to determinate the copper content in the waste slag depending on its chemical composition, following values of R-2 were obtained: 0.09 and 0.999. These results indicate that the ANN is more appropriate for solving the nonlinear problem, of this technological process at the industrial level, described in this paper.