Regression & Modeling · TOOL GUIDE

Gradient Boosting

Open Quantly ↗

Understand the method. Use it with confidence.

Boosted tree model for regression or classification with conformal uncertainty support. Quantly runs the method against your active analysis context and returns structured results for review and interpretation.

01

When to use it

Use this tool to quantify relationships, explain variation, or predict an outcome from one or more features.
Use it only when the selected variables and study design match the method's requirements.

02

How to use it

  1. Open the tool from the Quantly Tools panel.
  2. Confirm the active dataset and analysis session.
  3. Set Target Variable (Required): Outcome column to predict.
  4. Set Feature Columns (Required): Predictor columns used by the model.
  5. Set Categorical Encoding (Optional): How to encode categorical predictors before fitting.
  6. Set Calibration Holdout Share (Optional): Fraction of complete rows reserved for the held-out conformal calibration split.
  7. Set Boosting Stages (Optional): Number of boosting stages to fit.
  8. Set Learning Rate (Optional): Shrinkage applied to each boosting stage.
  9. Set Tree Depth (Optional): Depth of the weak learner trees.
  10. Review optional or advanced settings and keep defaults unless your design requires a change.
  11. Run the tool, inspect warnings and diagnostics, then save or export the result with its assumptions.
03

How to interpret results

Review coefficients or feature effects, uncertainty, validation metrics, residual diagnostics, and leakage risk. Association is not causation, and performance should be evaluated on data not used for fitting.