Forecasting Accuracy through Machine Learning in Supply Chain Management

Abstract

Purpose: The use of machine learning (ML) techniques in economic and financial forecasting has gained significant attention due to their potential to improve the accuracy and robustness of predictions. This paper explores the application of various ML algorithms such as support vector machines,  random forests, and deep learning models in forecasting economic variables, financial market trends, and macroeconomic indicators.

Methodology: We assess the forecasting accuracy of these models  relative to traditional econometric approaches, including ARIMA and VAR models.

Findings: The analysis  reveals that ML techniques, particularly deep learning, outperform classical methods in terms of predictive accuracy, especially in complex, nonlinear environments. We also discuss challenges associated with model interpretability, overfitting, and data quality, providing insights into how these limitations can be addressed.

Unique Contribution to Theory, Practice and Policy: The findings contribute to a deeper understanding of how advanced machine learning can enhance forecasting methodologies, with implications for both theoretical modeling and practical applications in economic policy, risk management, and financial decision-making.