Three Kinds of Learning

Machine learning comes in three main varieties:

• Supervised learning — the model learns from labeled examples. This powers spam filters, medical diagnosis, and image recognition
• Unsupervised learning — the model finds patterns in unlabeled data, discovering natural groupings without being told what to look for
• Reinforcement learning — the model learns by trial and error with rewards. This is how AlphaGo mastered Go and how robots learn to walk

Most real-world ML uses supervised learning, because labeled data is the most direct way to tell a model what you want.