AI-Based Forecasting Models for Macroeconomic Indicators
Keywords:
AI forecasting, macroeconomic indicators, XGBoost, ARIMA, LSTM, hybrid modelsAbstract
This study explores the performance of AI-based forecasting models in order to predict key macroeconomic variables, including the GDP growth rate, inflation rate, unemployment rate, interest rates and exchange rates. The study compares the predicting accuracy of their models through a set of machine learning models, including ARIMA, XGBoost, LSTM, and hybrid models. The results show that XGBoost has a great predictive potential in macroeconomic forecasting since it keeps outperforming other models with the lowest error indicators. The hybrid model, which is a combination of XGBoost and ARIMA with significant improvement in GDP and inflation rate forecasts is more accurate. Moreover, because it is evidenced by the great performance of the hybrid model in the diverse times of the day and economic circumstances, the analysis of these models underlines the importance of model selection in correct forecasting. The study also compares the envisaged and actual numbers of each indicator to indicate the resilience of the models in predicting volatile economic status. The findings point to the predictive power of AI-driven methods and their utility to the economic researcher and policy makers.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Asif Mehmood, Nida Farooq (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.
