Bankruptcy Prediction via Hybrid Resampling and Stacking Ensemble Techniques with Explainable Artificial Intelligence (XAI)-Driven Analysis

Not specified in the provided content

ensemble learningbankruptcy predictionexplainable AIimbalanced data

2608.20343

Builder Relevance

80%

2h ago

Abstract

This study presents a bankruptcy prediction framework that enhances minority-class detection in imbalanced financial data using advanced machine learning techniques.

Reality Card

Core Claim

The GRU model with SMOTE-ENN achieved the best predictive balance for bankruptcy prediction, with a recall of 0.8627 and ROC-AUC of 0.9431.

Method / Result

The hybrid stacking ensemble of SMOTE-ENN with (GB+XGB+HGB+LGBM+AB)+LSTM provided the strongest compromise between sensitivity and specificity.

Limitations

The study relies on specific datasets and resampling techniques that may not generalize across different financial contexts.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.

← Back to all papers