Theoretical bases of the application of various-color graphs in the solution of intellectual chemical tasks

Abstract

The paradigm of using artificial neural networks (ANN) for solving intellectual problems of chemistry and chemical technology is considered: classification, identification, design, modeling, optimization, and others. Using the example of studying the applicability of colored graphs in the neural network analysis of chemical structures at the site «structure-property-application» relationship, the possibility of identifying chemical structures when creating actual substances is shown. Artificial neural network learning to identify graphs is shown. The results obtained are mathematical software that allows solving creative problems and creating decision rules when choosing chemical-technological systems formalized in terms of graph theory and intended to support decision-making.