Comment on page
Feature Types
Discrete
- Numbers
- Categorical
- Categorical data are variables that contain label values rather than numeric values.
The number of possible values is often limited to a fixed set.
- labels, usually discrete values such as gender, country of origin, marital status, high-school graduate
Continuous (the opposite of discrete): real-number values, measured on a continuous scale: height, weight.
In order to compute a regression, categorical predictors must be re-expressed as numeric: some form of indicator variables (0/1) with a separate indicator for each level of the factor.
Discrete with many values are often treated as continuous, i.e. zone numbers - > binary