Week 9: Feature Engineering

Learning Objectives

• 09A I can apply preprocessing steps — scaling and one-hot encoding — and explain why each is needed before modeling.
• 09B I can build a scikit-learn pipeline that chains preprocessing and modeling steps into a single reproducible workflow.
• 09C I can identify how missing data arise (MCAR, MAR, MNAR) and choose appropriate imputation strategies for each case.

Perspectival Reading

Reading: TBD

Reflection Questions

1. Feature engineering requires domain knowledge — whose knowledge counts, and who is excluded from this process?
2. Imputation fills in missing values with estimates. What assumptions does a chosen strategy make about why data is missing?
3. When you encode a variable like gender or ethnicity, what are you doing to how the model treats those groups?

Materials

• Monday class
• Wednesday class