The solution of certain socio-economic problems related to the agricultural sector is becoming relevant in the light of population growth and changes in the environmental situation both in the world and in the Central Asian region. Long-term planning and forecasting of the dynamics of processes in agriculture is becoming the cornerstone of achieving food security and maintaining social stability in the country. In this paper, it is planned to consider the use of computer technologies and developments in the field of modeling for the creation, training and application of deep machine learning with a teacher to predict the volume of production of the main types of grain crops