Model Configuration
Decision Boundary Visualization
Tree Structure
Help
Decision Trees
- How it works: Recursively split data based on features to create homogeneous groups
- Gini Impurity: Measures probability of incorrect classification (lower is better)
- Information Gain: Reduction in entropy after a split (higher is better)
- Max Depth: Limits tree depth to prevent overfitting
Interpreting Results
- Decision Boundaries: Rectangular regions where tree makes predictions
- Tree Structure: Shows splits and leaf nodes with their predictions
- Training vs Test: Large gap indicates overfitting
- Min Samples: Higher values create smoother boundaries