Classification models map input data to an output prediction of a categorical class label, as opposed to regression models that predict numerical values. The distinction is most easily conveyed through examples:
- Classification models can assign labels such as true/false, low/medium/high risk, or which animal species.
- Regression models can predict output such as housing prices, long jump distances, or number of home runs hit.
I will use the prediction nomenclature described earlier in the chapter for the entire classification section. Make sure that you are familiar with the nomenclature before reading further.
The rest of this chapter will cover some common methods used for prediction. The following is a group of plots comparing different prediction methods and how they map input data to target variables:
