Labeled data includes a feature, or label, that identifies the target value or category a machine-learning model should predict. Whether a feature counts as a label depends on the problem being solved. section 2.6 section 2.5
Labels commonly take two forms:
- Numerical labels are numbers, such as an animal's weight or a house price; models that predict them use regression. section 2.7 section 3.37
- Categorical labels are finite categories or states, such as dog/cat or spam/ham; models that predict them use classification. section 2.6 section 2.7
Recognizing a new pet
A dataset of dog and cat images carries dog or cat labels. A supervised model remembers those examples, formulates rules, and predicts the label of a new unlabeled image. section 2.7
Figure 2.1 Labeled data is data that comes with a tag, or label. That label can be a type or a number. Unlabeled data is data that comes with no tag. The dataset on the left is labeled, and the label is the type of pet (dog/cat). The dataset in the middle is also labeled, and the label is the weight of the pet (in pounds). The dataset on the right is unlabeled.
Compare labeled datasets containing pet types or weights with an unlabeled dataset containing images alone. section 2.6







