Abstract
In the context of global climate change, the sustainable development of agricultural systems is becoming one of the priority tasks of modern agricultural science and practice. Uneven precipitation, rising average annual temperatures and more frequent extreme weather events increase the risks of reduced yields and economic instability in agriculture. This paper examined the application of artificial intelligence methods for forecasting the sustainability of agricultural systems at the regional level. The aim of the study was to develop approaches to forecasting the productivity and adaptive potential of agricultural crops using neural networks and machine learning algorithms. The materials and methods of the study included the use of long-term statistical data on the yield of major crops, climatic indicators, and economic parameters of agriculture in the Kyrgyz Republic. Correlation-regression modelling, artificial neural networks, and clustering algorithms were used for the analysis. The results of the study showed that the use of intelligent algorithms can increase the accuracy of yield forecasts by 12-15% compared to traditional methods, as well as identify key climatic and economic factors that determine the sustainability of agricultural systems. The scientific novelty of the work lies in the integration of artificial intelligence methods with agroecological zoning to build adaptive models of sustainable development. The practical significance of the research lies in the possibility of applying the developed models in strategic planning, the formation of regional food security programmes, and risk management in the agricultural economy