Cross-Market Bankruptcy Prediction: An Interpretable Ensemble Learning Framework Using SHAP Analysis

Puntos clave

• Ensemble learning demonstrates effective bankruptcy prediction across various markets, providing insights into financial health.
• Using SHAP analysis, the model reveals the impact of key variables on bankruptcy likelihood, enhancing understanding of financial indicators.
• The approach integrates various machine learning methods to improve predictive accuracy and model interpretability in bankruptcy scenarios.
• Findings support the need for interpretable models in finance, as understanding the prediction factors can guide better decision-making.