The Three Ways Machines Learn

Machine learning splits into three broad paradigms, based on what kind of experience the system learns from: supervised learning (learning from labeled examples, where the right answer is already known), unsupervised learning (finding structure in data with no labels at all), and reinforcement learning (learning through trial and error, guided by rewards and penalties).

Supervised learning is the most common starting point: show the system thousands of emails already labeled 'spam' or 'not spam,' and it learns to predict the label on a brand-new email. The 'supervision' is the labels — someone already did the work of saying what the right answer looks like.

Unsupervised learning has no labels to lean on — instead it looks for patterns or groupings that exist in the data on their own, like automatically clustering customers into groups with similar buying habits, without anyone pre-labeling which customer belongs to which group.

Reinforcement learning is different from both: there's no fixed dataset of labeled examples at all. Instead, an agent takes actions in an environment, gets a reward or penalty based on the outcome, and gradually learns which actions tend to lead to better rewards — the same basic loop a game-playing AI or a robot learning to walk actually uses.

What distinguishes supervised learning from unsupervised learning?

Learning through trial and error, guided by rewards and penalties for actions taken in an environment, is called ____.

Unsupervised learning relies on a dataset where every example already has the correct label attached.