Machine Learning Basics·Machine learning from data to a responsible model

What machine learning is

Separate learning from rules, and recognise when a model is the wrong tool.

Intermediate 12 min read Lesson 1 of 300 Challenge included

What you will be able to do

Supervised learning uses labelled examples.
Unsupervised learning finds structure without labels.
Start with the simplest baseline.
If a rule is reliable, write the rule.

In plain English

Machine learning fits a function from examples instead of hand-written rules. It is useful when the pattern is real, data is representative and mistakes are tolerable. It is not a replacement for understanding the problem.

framing.pyPython

# Supervised: predict a number or a class from labelled rows.
# Unsupervised: group similar rows without labels.
# Start with a baseline: mean, majority class or a simple rule.
baseline = 0.78
print(f'Baseline accuracy: {baseline:.0%}')

Your turn

Name the task

Easy

Given rows of house features and sale prices, is this regression or classification?

PYTHON · starter

task =

Check your understanding before moving on.

Knowledge checkQuestion 1

What makes learning "supervised"?

Select an answer to continue

Study tools

Make it yours.

Keep a private note for this lesson, then turn the key ideas into flashcards you can revisit.

Sign in to save notes and flashcards.Both are private to your account and stay with your progress.

Finished reading?

Lock in what you just learned.

Mark this lesson complete to save your place and move one step closer to the certificate.

Next: Data and features

Lesson 1 of 300 · 0% of course