Supervised learning
Learns from examples containing known targets.
- Classification
- Regression
- Forecasting
- Ranking
Classify realistic machine learning problems, uncover essential clues, build confident reasoning, earn achievements, and master supervised, unsupervised, and reinforcement learning decisions through interactive challenges.
Read the scenario card. Drag it onto a learning zone, or select a zone with your keyboard. Use clues carefully because each clue reduces the available score.
Configure a mode, then start the game.
Keyboard: press 1 for supervised, 2 for unsupervised, or 3 for reinforcement. Press H for a clue.
Learns from examples containing known targets.
Discovers patterns without supplied target labels.
Learns actions through rewards and penalties.
| # | Scenario | Your choice | Correct type | Result | Time | Confidence |
|---|---|---|---|---|---|---|
| Complete a round to build your review. | ||||||
Look for historical examples containing known target values or labels.
Look for discovery, grouping, compression, or unusual-pattern detection without reliable labels.
Find an agent choosing actions, receiving rewards, and influencing future states.
Yes. A system may predict outcomes, discover segments, and optimize decisions separately.
The clue reveals a decisive signal, but the question earns fewer points.
It compares certainty with correctness and highlights possible overconfidence.
Scores, experience, badges, settings, and custom scenarios use browser storage.
Yes. Use the custom scenario builder to add local question cards.
You begin with three lives and lose one after each incorrect classification.
Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.