How Machines Learn from Examples
Welcome! Most software follows rules that a person wrote down. In machine learning, you show the computer many examples and it finds the rules by itself. Here, you will watch a computer learn how the size of a house is linked to its price. You need no maths and no code.
New words
- Machine learning (ML)
- Computers find the rules by themselves, using many examples.
- Traditional programming
- A person writes every rule by hand for the computer to follow.
- Data
- The facts and numbers we give a computer, like house sizes and prices.
- Model
- The rule that the computer learned from examples.
- Error
- The gap between a guess and the real answer.
- Learning rate (step size)
- How big each small change is in one round of learning.
- Gradient descent
- A method that changes the rule in small steps to make the error smaller.
What you will learn
- Two ways to teach a computer
- An example: guessing the price of a house
- How does the computer find the line?
- Try it: teach the line
- Four ideas people often get wrong
- Where you meet machine learning every day
- Explain it to a friend
- Key takeaways
10 quiz questions at the end.
The full lesson, with interactive demos and the quiz, opens after you sign in with Google. It is free.