This project focuses on predicting the stage of liver cirrhosis using clinical patient data. It explores multiclass classification techniques by comparing Random Forest and XGBoost models. Key features include:
- ✅ Model comparison between Random Forest and XGBoost
- ⚙️ Hyperparameter tuning for performance optimization
- 🔄 SMOTE applied to address class imbalance
📊 The project aims to provide accurate classification of disease progression to support medical decision-making.
🧠 Note: To reduce the risk of late diagnoses (e.g., predicting Stage 2 when the actual stage is Stage 3), the model prioritizes recall for critical stages such as Stage 3 and Stage 4.