Neural network based on the negatrons

Artificial neural networks are widely used for building intelligent control systems, pattern recognition, prediction, optimization, diagnosis, clustering, associative memory and so on. Today there are many different options for implementing artificial neurons. Complex physical models of artificial neurons can most accurately reflect all processes functioning of biological neurons, however, they contain a large number of elements and the creation of large arrays of neurons is a complicated technical task. An alternative approach is to create the simplest possible hardware implementations of artificial neurons with preservation of the main functions of the neuron. Promising element base for creating these neurons are functional electronic devices, including R-, L-, C-negatron which use will provide the circuitry simplicity, high reliability, economy, technology, small size and weight [1].