Worksheets › Class 6 › Introduction to AI and Everyday Examples › Cheat sheet
Class 6 Introduction to AI and Everyday Examples: cheat sheet
CBSE Computational Thinking and AI. AI lets machines do tasks that need thinking, like recognising and predicting, by learning from data. Automation only repeats a fixed rule.
Key ideas
- Artificial Intelligence (AI) is the ability of machines to do tasks that usually need human thinking.
- AI learns from data and can improve; automation follows the same fixed rule every time and does not learn.
- Supervised learning learns from examples with the right answers (labels), such as photos marked “cat” or “dog”.
- Unsupervised learning finds groups in data that has no labels; reinforcement learning learns by trial and error with rewards.
- Machines are better at fast, repetitive work; people are better at feelings, common sense and imagination.
- AI can make mistakes, so we test it on new examples and check its answers.
Remember
| Accuracy | correct ÷ number tested × 100% |
| Example | 21 right out of 25 → 84% |
| Wrong answers | number tested − correct |
| Reward score | points for good moves × good moves − penalty × bad moves |
| AI or automation? | maps app predicting traffic: AI; street light on at 6 p.m.: automation |
Watch out
- Calling every machine AI: an automatic door with a fixed rule is only automation.
- Dividing by the number wrong instead of the number tested when finding accuracy.
- Mixing up supervised (labels given) and unsupervised (no labels) learning.
Worked example
(a) A robot learns a game by trial and error. It gets +9 points for a good move and loses 5 points for a bad move. In 27 moves, 21 were good. What is its score? (b) A model is tested on 25 pictures of fruit it has never seen. It labels 21 of them correctly. What is its accuracy?
Working: (a) 9 × 21 − 5 × 6; (b) 21 ÷ 25 × 100
Answer: (a) 159; (b) 84%
Practise Introduction to AI and Everyday Examples
Revise with this page, then practise: every Introduction to AI and Everyday Examples sheet in the app is new, with an answer key and working.
Coming soon for Android and iPhone.