4.3 Supervised Learning
Supervised learning = learning from labelled examples. Like studying with an answer key.
Chapter 4: Introduction to Machine Learning (ML)
In supervised learning, every training example comes with the correct answer (a label). Photo of a dog → label "dog". House details → label "sold for $300k".
The model makes a guess, compares it to the label, and adjusts. Over thousands of examples it learns the mapping from input to answer. Then it can predict labels for new, unlabelled examples.
It's the most common type of ML. Spam filters, price predictions, medical image classifiers and face recognition are all usually supervised. The catch: labelling data takes a lot of human time and money.
Key takeaways
Learns from labelled examples (input + correct answer)
Most common type of ML
Labelling data is expensive and slow