АЛГОРИТМЫ ЦИФРОВОЙ ОБРАБОТКИ ИНФОРМАЦИИ НА ОСНОВЕ НЕЙРОННЫХ СЕТЕЙ ДЛЯ РАСПОЗНАВАНИЯ ШИРОКОГО КЛАССА ХИМИЧЕСКИХ ВЕЩЕСТВ
In the development of algorithms for digital processing of information for detecting substances and gas mixtures according to their smell, obtained with the "electronic nose", it is proposed to use a quickly trained neural networks for signal processing. We demonstrate that classical quadratic forms are not able to solve the problem of recognizing high- dimensional images. The "deep" Galushkin-Hinton neural networks can solve the problem of high-dimensional image recognition, but their training has exponential computational complexity. It is technically impossible to train and retrain a "deep" neural network rapidly. For mobile "e-nose" systems, we proposed to employ a number of "wide" neural networks trained in accordance with state standards requirements (GOST R 52633.5-2011). This standardized learning algorithm has a linear computational complexity, i.e. for each new smell image, a time of about 0.3 seconds is sufficient for creating and training a new neural network with 2024 inputs and 256 outputs. This leads to the possibility of the rapid training of the artificial intelligence "e-nose" and a gradual expansion of its database consisting of 10 000 or more trained artificial neural networks. The article proposes the development of algorithms for processing data obtained with the use of self-learning intelligent neural networks.