AI-ASSISTED ANALYSIS OF TOURIST FLOWS AND THEIR IMPACT ON THE ECONOMY OF SAMARKAND

Samarkand is one of Uzbekistan's main cultural tourism destinations, and tourism has become an important driver of local employment, services and investment. Traditional statistics on visitor numbers are published with delays and at coarse spatial and temporal resolution, which limits their usefulness for planning. Objective: This study examines how artificial intelligence (AI) methods can be used to analyse and forecast tourist flows in Samarkand and to estimate their contribution to the city's economy. Methods: We combine official accommodation and border statistics, mobile and online-platform indicators, and card-transaction data. Machine-learning models (seasonal ARIMA, gradient boosting and an LSTM network) are compared for short-term forecasting, and an input-output framework is used to translate forecast visitor spending into output, income and employment effects. Results: [Insert main quantitative findings: best model and its error, peak-season share of arrivals, estimated spending per visitor, total and multiplier effects.] Conclusions: AI-based analytics can provide more timely and granular information about tourist behaviour than conventional statistics, and when linked to economic impact models they support evidence-based decisions on capacity, pricing, infrastructure and seasonality management.