A table lists houses with the columns size, rooms, neighbourhood and price. You want to predict the price. Which one is the label?
Options
- Size, because it is the most important column
- Neighbourhood, because it is text
- The whole table
- Price, the value you want to predict (Correct option)
Answer
Price, the value you want to predict
Why
The label (or target) is the value the model must predict. The other columns are features: the inputs it uses to do so.
Features can be numbers or categories. A text column such as neighbourhood is not a label, it just needs encoding first.
Which column is the label depends on the question you ask, not on the type of the column.
Remember
Label = what you predict; features = what you predict from.