AI Solutions for Digital Diagnostics of Grain Crop Diseases (Based on the Example of Pyrenophora teres in Winter Barley).

The purpose of this research is to justify the feasibility of using digital intelligent technologies in forecasting the development of net blotch in winter barley. The developed AI solution is a binary decision tree that can predict scenarios of net blotch development: depressive, moderate, and epiphytotic development. To configure the algorithm parameters, we carried out field and laboratory experiments at the Federal Scientific Center for Biological Plant Protection from 2021 to 2023. The preparation of data involved several stages, including setting up of field plots to create an artificial infection background as well as the preparation of an inoculum, sowing of highly susceptible and resistant winter barley varieties, and artificial inoculation. The selected input factors included the observed degree of leaf damage, type of variety resistance, vegetation phase at the time of primary infection, and average relative air humidity during the vegetation phase of infection. The total sample size was 144 observations. The trained model has demonstrated a high classification accuracy on both the training and test datasets at an accuracy rate of more than 96%. Based on the statistical estimate of the significance of the factors influencing the development of net blotch in barley, it is shown that the most influential factor is the current degree of leaf infection (74.3%), followed by the average relative air humidity (11.9%), the resistance of the variety to the disease (10.4%), and the development stage during which infection occurred (3.4%). The proposed solution has a significant practical importance since it provides new opportunities for the diagnostic process of net blotch in winter barley, including high diagnostic rate, accuracy in forecast predictions, and applicability in field conditions. [ABSTRACT FROM AUTHOR]