0x00Lesson 1 of 11

Understand machine learning problems

Frame a problem as classification, regression or clustering, and name its examples, features and target.

18 min 6-question quiz 2 code exercises

By the end of this lesson you can

Distinguish supervised and unsupervised learning
Tell classification from regression
Decide when machine learning is the right tool at all

Ordinary programs follow rules a person wrote. Machine learning (ML) programs learn their rules from examples. Show a model thousands of houses with their prices, and it learns a function from house features to price; then it can price a house it has never seen.

Supervised learning - every training example has a known answer, the target (or label). Predicting a category is classification (spam or not, which disease); predicting a number is regression (price, delivery time).
Unsupervised learning - no targets. The model finds structure: clustering groups similar examples; dimensionality reduction compresses many features into a few.

Each example is described by features - the inputs, like rooms and area. Formally, a supervised model learns a function $\hat{y} = f(x)$ that maps features $x$ to a prediction $\hat{y}$ close to the true target $y$ .

Try it

Up to 15 XP

What kind of problem is it?

Sort each task.

0 of 6 sortedScore 0/0

“Flag card transactions as fraudulent or legitimate”
“Estimate how many minutes a delivery will take”
“Group customers by shopping behavior, with no predefined segments”
“Compute sales tax from a fixed tax table”
“Predict next month’s electricity demand in megawatts”
“Recognize which of 10 digits a handwritten image shows”

examples.py

1houses = [
2    {"rooms": 3, "area_m2": 82, "price": 275000},
3    {"rooms": 2, "area_m2": 55, "price": 190000},
4]
5features = [[house["rooms"], house["area_m2"]] for house in houses]
6targets = [house["price"] for house in houses]
7print(features, targets)

Output

[[3, 82], [2, 55]] [275000, 190000]

Key takeaways

Supervised learning learns $\hat{y} = f(x)$ from examples with known targets.
Classification predicts categories; regression predicts numbers; clustering finds groups without labels.
Define the real task, the target and what a useful result is before picking an algorithm.

Lesson quiz

6 questions · pass with 5 correct · up to 50 XP

Passing this quiz completes the lesson and keeps your streak going. Questions you miss come back in review sessions later.

Practice: write Python

Write Python in the editor and run it against sample inputs. Python runs locally in your browser using a WebAssembly runtime.

Exercise 1

Split features from the target

+25 XP

Read a CSV where the last column is the target. Print features: NAMES (comma-separated), target: NAME, examples: N, and task: classification if every target value is non-numeric, otherwise task: regression.

House prices
Spam

main.py

Loading editor…

Python runs in a sandboxed browser worker with a 60 second time limit. Its runtime loads from the Pyodide CDN; your code stays in this browser.

Exercise 2

Beat the baseline

+25 XP

Line 1 lists training labels; line 2 lists test labels (space-separated). A baseline that always predicts the most common training label (alphabetically first on ties) is the first thing any model must beat. Print baseline: LABEL and baseline accuracy: P% on the test labels (whole percent).

Imbalanced classes

main.py

Loading editor…

Python runs in a sandboxed browser worker with a 60 second time limit. Its runtime loads from the Pyodide CDN; your code stays in this browser.

Questions about this lesson

Stuck? Ask. Figured something out? Share it. Explaining is one of the best ways to learn.

Loading posts…