Predicting stock market crashes with machine learning: A review and methodological proposal

Abstract

This review paper examines the utilisation of machine learning techniques for predicting stock market crashes. It surveys existing methodologies, identifies common trends, and analyses strengths and weaknesses. A novel methodological framework is proposed, integrating ensemble learning, alternative data sources, and model interpretability to address limitations in current approaches. The proposed framework aims to enhance predictive accuracy, transparency, and actionable insights in financial forecasting. Future research directions include empirical validation, interdisciplinary collaboration, and the integration of emerging technologies. Continued research in leveraging machine learning for financial forecasting is vital for advancing risk management practices and fostering resilient financial systems.