Machine Learning

Corporate Bankruptcy Prediction Using Machine Learning

The Corporate Bankruptcy Prediction Using Machine Learning system is a comprehensive machine learning solution that predicts corporate bankruptcy using financial ratios and company metrics. This project is ideal for B.Tech, MCA, CSE, and engineering students as their final year project. It implements LightGBM and XGBoost with SMOTE for class imbalance, achieving 96.7% accuracy and 0.991 ROC-AUC.

The system utilizes the Taiwan Economic Journal (TEJ) Corporate Bankruptcy Dataset containing 6,819 company records with 95 financial features, including profitability ratios, leverage indicators, and market metrics. It provides early warning capabilities for financial risk management, enabling investors, creditors, and regulators to identify at-risk companies proactively. Feature importance analysis reveals that profitability ratios, particularly ROA, Net Value Per Share, and Operating Gross Margin, are the strongest predictors of bankruptcy.

• Multi-Model Bankruptcy Prediction
• XGBoost (96.7% Accuracy)
• LightGBM (95.3% Accuracy)
• Leak-free SMOTE Oversampling
• Feature Selection (SelectKBest)
• Interactive Visualizations
• Real-time Bankruptcy Predictions
• Model Performance Comparison
• Feature Importance Analysis
• Flask Web Application