Fundamentals
Label
A label is the target value a supervised model learns to predict.
How it works
Training compares a prediction with its label to calculate error. Labels may be classes, numbers, sequences, or structured objects.
A concrete example
Illustrative: for an email, the label is spam = 1 while 42 token counts are features.
Do not confuse it with Feature
a label is the answer; a feature is evidence used to infer it.