Simple Software for Training Artificial Neural Network Models Based on the Block Coding Approach

The paper focuses on the development of a software program for creating and training ANN (artificial neural network) models. This program allows users to create their own ANN models without requiring programming knowledge or experience with specialized libraries. The customization of ANN models is done using a block-coding approach, i.e., using interactive graphical scene tools without writing code in textual form. The paper describes the architecture, graphical user interface and functional features of the developed software. The paper concludes with the results of testing the software on an example of training the ANN model for predicting the shielding effectiveness of planar metal electromagnetic shields. It was shown that the average deviation between the results obtained by the trained ANN model and using the domestic electromagnetic simulation system does not exceed 2 dB. The results confirm the validity of the program and the acceptable accuracy of the ANN model created with its help.