PROSPECTS FOR THE USE OF NEURAL NETWORKS IN CROP BREEDING

Issue № 5, 2025, article № 10, pages 113-120

Section:

Language: Russian

Original language title: ПЕРСПЕКТИВЫ ИСПОЛЬЗОВАНИЯ НЕЙРОСЕТЕЙ В СЕЛЕКЦИИ СЕЛЬСКОХОЗЯЙСТВЕННЫХ КУЛЬТУР

Keywords: SELECTION, FOOD INDUSTRY, FOOD SECURITY, NEURAL NETWORKS, ARTIFICIAL INTELLIGENCE, MATHEMATICAL METHOD

Abstract: The article discusses the prospects of using neural networks in the selection of crops for the food industry. The important role of the agro-industrial complex (AIC) in the Russian economy, especially in ensuring food security, is emphasized. It is noted that breeding is a key branch of the agro-industrial complex, aimed at creating and improving crop varieties with specified traits, such as increasing yields, improving product quality and creating resistant varieties. The article emphasizes that the process of developing a new variety can take considerable time and requires funding for experimental sites and researchers. Therefore, the possibility of using mathematical methods and neural networks to speed up the breeding process is discussed. The historical experience of using mathematical models in breeding and their disadvantage as a tool are considered. Difficulties of use. Neural networks, unlike mathematical models, do not require explicit definition of mathematical equations and are able to approximate complex functions. The paper summarizes the capabilities of neural networks, and in which of the areas of selection they can be useful. The difficulties of implementation of neural networks approaches for wide application are emphasized. Such as, creation of modern infrastructure for data collection and processing, as well as training of specialists with experience in machine learning, genetics and agriculture. Difficulties in implementing neural network tools are described, due to mistrust or lack of understanding of new tools and techniques among agrarians. It is noted that, despite the prevalence of neural networks in commercial activities, their application in the field of agrocomplex and food industry is still rare.

Authors: Naumova Olga Alexandrovna, Naumov Denis Valerievich