Application of machine learning in the organization of adaptive-landscape farming systems

The article explores the possibilities of using artificial intelligence tools in organizing ecological landscape farming systems. It is noted that with the help of learning algorithms, agricultural producers can optimize many production processes, increase the productivity of land and the quality of the resulting products, as well as reduce costs and costs. It has been shown that the mass implementation of machine learning can increase the share of gross value added in the next 5 years by 25% in crop production, and up to 14% in livestock production under an optimistic scenario of development; in the most probable scenario, the indicators will be two times lower, and in a pessimistic scenario - an increase will occur by no more than 3.8% in the crop industry, and up to 0.4% in the livestock industry. Since adaptive landscape agriculture, based on taking into account the characteristics of the relief, climate, agrofacies, must take into account a large number of parameters, such as assessment of the condition of the soil and plants, sown areas, frequency of their cultivation, the amount of applied mineral and organic fertilizers, treatment with herbicides and insecticides, etc. ., a prototype of an information system was developed that allows, based on predictive analysis, to select the most optimal solution for organizing crop rotations in order to manage farming systems. The article shows the possibility of using computer vision in recognizing cartographic material and establishing the type of agricultural landscape to obtain highly productive crops. Intellectual analysis models were built based on incoming features. To work with the proposed product, there are no specialized requirements for personnel qualifications, and can even be used by ordinary employees of both large agricultural holdings, representatives of municipal and state authorities in the agricultural sector, and employees of small farms due to its simplicity and intuitive interface.