ANALYSIS AND COMPARISON OF SELECTED DATA SCIENCE MODELS IN THE CONTEXT OF THEIR APPLICATION TO MACROECONOMIC FORECASTING

Abstract

This article presents a review and practical comparison of selected Data Science models that can be applied to modern macroeconomic forecasting. The study focuses on the use of these models under conditions of increasing economic instability and structural changes. Particular attention is paid to analyzing the potential for combining classical econometric approaches with machine learning methods in order to improve the accuracy of macroeconomic forecasts. The empirical analysis is based on a real time series of nominal gross domestic product constructed using quarterly statistical data. Within the framework of the study, three forecasting models were developed and analyzed: the classical Autoregressive Integrated Moving Average (ARIMA) model, the ensemble Gradient Boosting Regressor (GBR), and the Multilayer Perceptron (MLP) neural network. The original dataset was divided into training and test samples, which allowed for an objective evaluation of forecasting performance on out-of-sample data. Standard forecast accuracy measures, including Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE), were used for comparative analysis of the results. Special attention is given to the investigation of boosting approaches as tools for enhancing the predictive capabilities of classical time series models. In particular, the potential for improving ARIMA-based forecasts through integration with machine learning techniques is examined. Based on the obtained results, the most promising model type and specification were identified, demonstrating higher forecasting accuracy compared to traditional econometric approaches and significant potential for further improvement. The practical significance of the study lies in the development of a methodological framework for applying Data Science tools in macroeconomic analysis and forecasting. This is especially relevant in the context of strategic planning and the economic recovery of Ukraine in the medium- and long-term perspective.