🧠 How Machines Learn: Models of Training ML

📚 Syllabus Dot Point

Explore models of training ML
Including: supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning

In the previous lesson, you learned that machine learning is about computers discovering patterns from data rather than following human-written rules. But how exactly does the learning happen? How does a computer go from knowing nothing to being able to recognise faces, recommend songs, or detect fraud?

The answer depends on what kind of data you have and what kind of problem you are trying to solve. There are four main models of training, each suited to different situations:

Supervised learning is like a teacher grading practice tests. You give the computer examples with the correct answers, and it learns the patterns that connect questions to answers.

Unsupervised learning is like sorting a pile of mixed laundry without being told which pile is which. The computer finds natural groupings in data without being told what the groups are.

Semi-supervised learning is a mix of both. You label a small portion of the data (expensive and time-consuming), and the computer uses those labels plus the larger unlabelled dataset to learn more efficiently.

Reinforcement learning is like training a puppy. The computer tries actions, receives rewards for good outcomes and penalties for bad ones, and gradually learns the best strategy through trial and error.