DS303: Feature Engineering & Data Preparation

SFIA Reference: DENG Level 4

Data Scientist (Grade 3)

Competency Description

Prepares data and engineers features to improve model performance and maintain data quality.

Learning Outcomes

• Apply encoding, scaling, and imputation techniques.
• Create domain-informed derived features.
• Collaborate with engineers on reusable transformation layers.

Evidence Requirements

Level 4:

• Delivered a transformed dataset for modelling.
• Used pipelines to automate preparation steps. Level 5:
• Developed reusable feature engineering code or patterns.
• Reviewed and optimised transformation pipelines.

Suggested Learning Activities

• Feature engineering lab using sklearn pipelines.
• Reflection on effective vs ineffective features in a project.
• Collaborate on feature store design.

Recommended Resources

• KnowHow: “Transforming with Purpose”
• Book: “Feature Engineering for Machine Learning” by Alice Zheng