Mathematical forecasting methods in a dynamic economy: integration of classical and АІ approaches in economic research

Abstract

The article examines modern methods of economic forecasting, their classification, advantages and limitations. Particular attention is paid to the application of artificial intelligence and machine learning methods in comparison with traditional quantitative approaches. Using the example of sales data at gas stations of Euro Smart Power LLC (Vinnytsia), a practical comparison of the effectiveness of various forecasting methods was conducted: the naive method, moving average, exponential smoothing, trend analysis, and intelligent forecasting using the Forecasting Pro program. The results of the study showed that hybrid approaches that combine artificial intelligence methods with classical statistical models demonstrate the highest accuracy, especially in the presence of seasonal fluctuations and other complex factors. The article emphasizes the importance of integrating quantitative and qualitative methods to improve the quality of forecasts in modern economic conditions.