Machine Learning for Kids Worksheet

AI for students lesson 2: How machines learn from examples · Set 18 · 40 questions

AI tip (Show, label, learn): Machine learning learns patterns from many labelled examples — a label is the right answer tag on each example. Good for training: many varied examples, correct labels, every type the model will meet, and separate new data for testing. Bad for training: very few examples, wrong or random labels, only one kind of example, blurry examples, testing on the training data.

1. 1.🧠 Machine learning is a way for computers to… (a) clean screens (b) charge faster (c) print pages (d) learn from data
2. 2.True or false: the message “Match practice moved to 5 pm” should be labelled “spam”. (a) True (b) False
3. 3.🔍 Which is the odd one out? (a) clear, sharp examples (b) copies of one photo (c) wrong labels (d) testing on training data
4. 4.📚 Good or bad for training a model: labels added at random? (a) good for training (b) bad for training
5. 5.📚 Usually, more good examples make a model… (a) change colour (b) forget everything (c) slower to switch on (d) better at its task
6. 6.True or false: the message “Send your password to claim a prize” should be labelled “not spam”. (a) False (b) True
7. 7.True or false: the message “Science project groups are on the board” should be labelled “not spam”. (a) False (b) True
8. 8.True or false: “clear, sharp examples” is good for training a model. (a) True (b) False
9. 9.True or false: “testing on training data” is bad for training a model. (a) True (b) False
10. 10.🔍 Which is the odd one out? (a) new data for testing (b) photos in many lights (c) clear, sharp examples (d) only one kind of example
11. 11.🔍 Which is the odd one out? (a) many varied examples (b) checking labels twice (c) very few examples (d) photos in many lights
12. 12.🏷️ You are labelling messages to train a spam filter. How should this one be labelled? “Your account is locked, pay ₹99 now” (a) not spam (b) spam
13. 13.🥭 A model saw only ripe yellow mangoes. A raw green mango arrives. What may happen? (a) it will shut down (b) it will turn yellow (c) it may get it wrong (d) it will be perfect
14. 14.Which of these is good for training a model? (a) testing on training data (b) only one kind of example (c) examples of every type (d) copies of one photo
15. 15.🐶 To teach a model to spot dogs in photos, what do you give it? (a) one single photo (b) many labelled photos (c) a song about dogs (d) a list of rules only
16. 16.Which of these is bad for training a model? (a) correct labels (b) many varied examples (c) new data for testing (d) testing on training data
17. 17.🏷️ What is a “label” in machine learning? (a) the right answer tag (b) the price (c) the font size (d) a sticker on a laptop
18. 18.Which of these is good for training a model? (a) correct labels (b) very few examples (c) wrong labels (d) blurry, unclear photos
19. 19.Which of these is good for training a model? (a) only one kind of example (b) very few examples (c) copies of one photo (d) new data for testing
20. 20.🔍 Which is the odd one out? (a) correct labels (b) new data for testing (c) wrong labels (d) many varied examples
21. 21.Which of these is bad for training a model? (a) clear, sharp examples (b) new data for testing (c) checking labels twice (d) only one kind of example
22. 22.📚 Good or bad for training a model: new data for testing? (a) bad for training (b) good for training
23. 23.📚 Good or bad for training a model: very few examples? (a) good for training (b) bad for training
24. 24.True or false: the message “Match practice moved to 5 pm” should be labelled “not spam”. (a) False (b) True
25. 25.Which of these is bad for training a model? (a) new data for testing (b) correct labels (c) checking labels twice (d) very few examples
26. 26.🔍 Which is the odd one out? (a) copies of one photo (b) labels added at random (c) photos in many lights (d) only one kind of example
27. 27.🔍 Which is the odd one out? (a) copies of one photo (b) blurry, unclear photos (c) very few examples (d) examples of every type
28. 28.📚 Good or bad for training a model: photos in many lights? (a) good for training (b) bad for training
29. 29.🏷️ You are labelling messages to train a spam filter. How should this one be labelled? “Science project groups are on the board” (a) not spam (b) spam
30. 30.True or false: the message “Your account is locked, pay ₹99 now” should be labelled “spam”. (a) False (b) True
31. 31.True or false: “examples of every type” is bad for training a model. (a) False (b) True
32. 32.Which of these is bad for training a model? (a) blurry, unclear photos (b) clear, sharp examples (c) new data for testing (d) correct labels
33. 33.🏷️ You are labelling messages to train a spam filter. How should this one be labelled? “Match practice moved to 5 pm” (a) not spam (b) spam
34. 34.True or false: “correct labels” is good for training a model. (a) False (b) True
35. 35.🔍 Which is the odd one out? (a) correct labels (b) only one kind of example (c) copies of one photo (d) wrong labels
36. 36.📚 Good or bad for training a model: only one kind of example? (a) good for training (b) bad for training
37. 37.Which of these is bad for training a model? (a) new data for testing (b) clear, sharp examples (c) correct labels (d) labels added at random
38. 38.📚 Good or bad for training a model: testing on training data? (a) bad for training (b) good for training
39. 39.True or false: the message “Please buy milk on the way home” should be labelled “not spam”. (a) True (b) False
40. 40.🏷️ You are labelling messages to train a spam filter. How should this one be labelled? “You won a free phone! Click now!” (a) not spam (b) spam

Free ai for students lessons, practice and worksheets at talentjr.in/ai-for-students