Realizations of the Artificial Neural Network for Process Modeling. Overview of Current Implementations

This work is intended to review the most typical realizations of Artificial Neural Networks (ANNs), implemented in a Feedforward Neural Network (FNN) as well as a Recurrent Neural Network (RNN). Essential differences in ANN architecture and basic operating principles are discussed. The problems of learning processes are presented in several cuts. The advantages of prediction using ANNs have been demonstrated in several popular fields such as adaptive educology, classification of medicine and biology, industry, etc.