Supervised Learning

In short

A type of Training where you have both the task at hand and the solution. Your goal is to find the right parameters that, given the task, will produce the solution you already know.

Like doing homework where you have the answer key at the back of the book. You practice solving problems and check your answers to improve.

In Machine Learning, there are different strategies for how a Model learns from Data. Supervised learning is the most intuitive one. A classic example: people want to buy or sell a house and the price can be a gray zone. You can use market history to build a model that, given the parameters of a house (size, location, rooms, etc.), predicts a good price for it. You have the inputs (house features) and the expected outputs (actual prices), so the model can learn the relationship between them.

Related

• Training - supervised learning is a type of training
• Unsupervised Learning - the other main approach
• Data - needs labeled data (inputs + expected outputs)