Module 2: Introduction to ML fundamentals

Introduction to Machine Learning Fundamentals

Slides

View slides in full screen

Outline

• Key ML terminology: model, training, features, target
• Decision trees as an intuitive starting point
• Train/Validation/Test splits
• Generalization, overfitting/underfitting
• The golden rule: never train on your test data
• Common data challenges:
  • Missing data
  • Mixed data types (numeric, categorical, binary, text)
  • Different feature scales
  • Outliers
  • Small datasets
  • Proxy variables
• Overview of a typical ML pipeline