←→ Select card
1 Classification
2 Regression
3 Clustering
4 Reduction
H Hint, Space Pause
Mistake review
Study incorrect placements and retry difficult cards.
| Algorithm | Your choice | Correct group | Explanation |
|---|---|---|---|
| No mistakes recorded. | |||
Achievements
Create a custom algorithm card
Saved custom cards
Session history
Stored locally in this browser.
| Date | Player | Mode | Difficulty | Score | Accuracy | Cards |
|---|
Export and share results
Download CSV or PDF reports, or copy a summary.
How the four groups differ
Classification
Classification predicts discrete labels. It uses labelled examples during training. Typical outputs include spam, disease, or image classes.
Regression
Regression predicts continuous numeric values. It learns relationships from labelled records. Common targets include prices, temperatures, and demand.
Clustering
Clustering discovers natural groups without target labels. It compares similarity among observations. Customer segmentation is a common application.
Dimensionality reduction
Dimensionality reduction compresses many features. It preserves useful structure where possible. It supports visualisation, denoising, and faster modelling.
Frequently asked questions
Why can one algorithm appear in different groups?
Some algorithm families support multiple prediction tasks. The card label identifies its intended variant. Explanations clarify accepted alternative uses.
Is logistic regression a regression algorithm?
Despite its name, logistic regression usually performs classification. It estimates class probabilities. The final output is commonly a category.
Are all reduction methods unsupervised?
No. PCA is usually unsupervised. Linear discriminant analysis can use labels for supervised reduction.
How are points awarded?
Correct answers add points. Fast responses and streaks add bonuses. Hints and mistakes reduce the score.
Where is progress saved?
Progress is stored in browser local storage. It stays on the current device. Clearing browser data removes saved progress.
Can teachers add algorithms?
Yes. Custom cards include names, categories, hints, uses, and explanations. They can form a dedicated practice pack.