AN INTELLIGENT MACHINE LEARNING BASED INFLATION FORECASTING

Abstract

This research study delves into the intricate dynamics of inflation forecasting, emphasizing the significance of machine learning techniques in contrast to traditional statistical approaches. It investigates the interplay between inflation and various economic variables such as interest rates, money supply, gold prices, and exchange rate across six diverse countries with different economic landscapes. By examining data spanning from January 1999 to December 2022, encompassing historical trends and economic shifts, this research explores the predictive capabilities of machine learning models in comparison to the traditional econometric models. Key findings reveal that no single model consistently outperforms others across all countries, underscoring the complexity and context-dependence of economic forecasting. The implications for model selection in inflation forecasting emphasize the importance of context-specific approaches, balancing complexity and interpretability, and continuous evaluation to adapt to evolving economic conditions. The study contributes to the ongoing discourse on hybrid methods for economic forecasting, suggesting a synthesis of machine learning and econometric models for enhanced accuracy and flexibility. Overall, this research offers valuable insights into the evolving landscape of inflation forecasting, providing policymakers, economists, and financial institutions with nuanced perspectives and methodologies for informed decision-making in an ever-changing global economic environment.

Keywords: Inflation Prediction, Macroeconomic Variables, Machine Learning, Traditional Models