An end-to-end automated machine learning pipeline that detects problem type, preprocesses data, trains multiple models, and selects the best one automatically.
- Automatic problem type detection (classification / regression)
- Dynamic preprocessing pipeline:
- Missing value handling
- Encoding categorical features
- Scaling numerical features
- Model training with cross-validation
- Multiple model candidates
- Best model selection
- visualize the metrics ( confiusion matrix)
- Clean modular architecture