Skip to content

10. Compute Feature Table

In this tutorial, you will learn how to generate a Feature Table from a feature list created in previous tutorials. Once the table is generated, we will train a LightGBM model on the new training data.

Step 1: Compute Feature Table

Navigate to the Feature List Catalog under the Experiment section of the menu.

For the 'Suggested SHAP selection (1+2) after adjusted ideation' feature list, follow these steps:

  1. Click Compute Icon. Name

  2. Select the Observation Table Applications up to Sept 2024 with Loan Defaults. Name

  3. Confirm the computation by clicking Compute Button. Name

Step 2: Review Feature Table

  1. Navigate to the Feature Table Catalog under the Experiment section of the menu. Name
  2. Click on the Suggested SHAP selection (1+1) after adjusted ideation - Applications up to Sept 2024 with Loan Defaults table. Name
  3. Open the 'Preview' tab to examine the table and select individual columns by clicking on column filter. Name Name

Step 3: Train a Model

Download the files lgbm_credit_default_ui_tutorials.ipynb and modeling_script.py here. Use the notebook to evaluate the accuracy. Name

Step 4: Iterate

The AUC of a model trained on applications up to March 2024 and validated on the next 6 months is 0.80.

Now, it's your turn!

Suggestions:

  • Train the model with the other feature lists in the catalog and create new ones.
  • Refine the ideation process, such as additional filters for BUREAU table.
  • Use the SDK to customize features.
  • Use the holdout observation table to further validate the model.