Label (Machine Learning)
A label is the correct answer attached to a training example in supervised learning, such as "cat" for a picture of a cat or the sale price of a house.
What is Label (Machine Learning)?
Labels tell the model what it should predict. They may come from human annotators, existing records or automated processes, and errors or inconsistent labelling directly reduce model quality.
How it works
During training the model's prediction for each example is compared with its label, and the difference drives the parameter updates.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions