Select Your Model
The Pareto Frontier
After training completes, the Pareto frontier chart displays all optimal models found during multi-objective optimization:
- X-axis: Model complexity (number of factor tables after consolidation)
- Y-axis: Predictive performance (cross-validation score—higher is better)
Each point represents a complete GBM configuration. Points on the frontier are Pareto-optimal: improving one objective (accuracy or simplicity) requires sacrificing the other.
Reading the Chart
Left vs. Right:
- Left side: fewer tables, more interpretable
- Right side: more tables, higher accuracy, still fully transparent
Making Your Selection
Consider your requirements:
| Use Case | Recommended Choice |
|---|---|
| Understanding key drivers (feature inference) | Simplest or Balanced |
| Operational pricing needing accuracy and transparency | Balanced or Most Accurate |
| Maximum predictive power | Most Accurate |
Click any point to select it.
After Selection
The system will:
- Fit the selected model configuration on the full training dataset
- Convert the GBM to factor tables
- Calculate test set performance (if test data was provided)
- Display factor tables, feature importance, and performance metrics