Applying the artificial intelligence neural network systems in achieving sustainable tourism development

Artificial intelligence has the capabilities to analyze large data sets, make decisions, and predict the stable behavior of various systems. Neural networks as one of the ways to implement artificial intelligence are a promising technology adaptive to real processes. It allows us to study multidimensional phenomena and work with incomplete data characterized by uncertainty and their uncertain mutual influence. The Kohonen neural network is applicable to finding patterns in data sets, identifying features, forecasting, modeling, and can be used to support the tourism sustainable development. The article uses indicators that affecting the tourism sustainable development on the territory of the Russian Federation for the period of 5 years. Environmental, economic, and social indicators of achieving the sustainable development goals of the Russian Federation, approved by the UN General Assembly in 2017, are also included for the analysis. The analysis encompasses 24 indicators. Using the method of self-organizing maps (one of the versions of Kohonen neural networks) based on an unsupervised network (unsupervised learning), the problem of clustering and projecting a multidimensional space into a two-dimensional one was solved. With the help of the Neural Network Toolbox for MATLAB software package, a graphical interpretation of the topology, distances, sample matches and input weight planes of the self-organizing map was given. As a result, correlated classes of indicators of sustainable development were identified, which make it possible to conduct further research to predict the tourism sustainable development and make informed management decisions.