Experiment for Creating a Neural Network with Weights Determined by the Potential of a Simulated Electrostatic Field
P. Sh. Geidarov · Scientific and Technical Information Processing · 2022
Abstract In this paper, based on the architecture of an artificial neural network using the metric recognition method, experiments are carried out to determine the weights and thresholds of the neural network with an electrostatic field parameter (potential) without any additional analytical calculations or learning algorithms. The simulation of the electrostatic field is implemented in the Builder C++ software environment, which evaluates the total potential of the electrostatic field at the points of the proposed model where potentiometer sensors are located. The same software module enables the creation of a neural network based on metrical recognition methods for which the weights of the first-layer neurons are determined based on the potentials of the simulated electrostatic field. The effectiveness of the resulting neural network is checked against the Modified National Institute of Standards and Technology (MNIST) database.