6.2.2 Learning Methods

ML models use two main types of methods for learning:

• Supervised (labelled) - dataset includes both the independent (predictors) and dependent (outcome) variables.
• Unsupervised (unlabelled) - only includes the input variables, so there is no direct objective, or “response” variable.
• Model choice can depend on several factors, e.g., data type, complexity, research question.

Start simple!