Features & labels

Every supervised dataset is a table: input columns are features, the column you predict is the label.

A feature is a measurable input (size, bedrooms, age). The label (a.k.a. target) is the answer you want back. Better features usually beat a fancier algorithm.