Build Your Own AI Project Worksheet
AI tip (Accuracy = correct ÷ total × 100): Steps: 1 choose a problem → 2 collect examples → 3 label them → 4 check and correct bad labels (clean the data) → 5 split into training and test sets → 6 train → 7 test with new examples → 8 measure accuracy → 9 improve the data and retest. Accuracy = correct ÷ total × 100. Wrong = total − correct. Expected right = accuracy ÷ 100 × total. Keep some examples apart for testing (for example one fifth); the rest are for training.
NameDateTime takenScore ___ / 40
- 1.📋 Planning an AI project: which comes first — fix it and test again or train the model? (a) fix it and test again (b) train the model
- 2.🧪 What is accuracy? (a) total × 2 (b) correct + total (c) correct ÷ total × 100 (d) total − correct
- 3.📋 Planning an AI project: which comes first — work out the accuracy or split train and test? (a) work out the accuracy (b) split train and test
- 4.🎯 Diya’s leaf identifier is 80% accurate. If it is tested on 60 new leaf photos, how many should it get right?
- 5.📋 Planning an AI project: which comes first — work out the accuracy or test on new photos? (a) work out the accuracy (b) test on new photos
- 6.📋 Planning an AI project: which comes first — split train and test or correct bad labels? (a) split train and test (b) correct bad labels
- 7.🎯 Aarav’s bird-song app is 70% accurate. If it is tested on 10 new bird songs, how many should it get right?
- 8.📋 Planning an AI project: which comes first — work out the accuracy or improve data, retest? (a) work out the accuracy (b) improve data, retest
- 9.⚖️ Two recycling sorter models were tested on the same 25 waste photos. Model A got 25 right and Model B got 20 right. How many percentage points more accurate is Model A?
- 10.🎯 Nisha’s fruit sorter was tested on 10 new fruit photos and got 6 right. What is its accuracy in percent?
- 11.📋 Planning an AI project: which comes first — correct bad labels or test on new photos? (a) correct bad labels (b) test on new photos
- 12.📋 Planning an AI project: which comes first — split train and test or recheck every label? (a) split train and test (b) recheck every label
- 13.❌ Kabir’s bird-song app was tested on 32 new bird songs and got 30 right. How many did it get wrong?
- 14.📋 Planning an AI project: which comes first — tag each photo by type or correct bad labels? (a) tag each photo by type (b) correct bad labels
- 15.🧪 Why must you test a model on examples it has NOT seen? (a) to make it faster (b) to make it bigger (c) to save paper (d) to see if it learned
- 16.🎯 Rohan’s spam spotter is 80% accurate. If it is tested on 30 new messages, how many should it get right?
- 17.🎯 Aarav’s handwriting reader is 80% accurate. If it is tested on 30 new handwritten letters, how many should it get right?
- 18.📋 Planning an AI project: which comes first — improve data, retest or work out the accuracy? (a) improve data, retest (b) work out the accuracy
- 19.⚖️ Two recycling sorter models were tested on the same 25 waste photos. Model A got 19 right and Model B got 18 right. How many percentage points more accurate is Model A?
- 20.🧪 Your fruit sorter mixes up oranges and lemons. What helps most? (a) fewer examples (b) a new colour screen (c) a louder speaker (d) more labelled examples
- 21.🎯 Diya’s bird-song app was tested on 25 new bird songs and got 15 right. What is its accuracy in percent?
- 22.📦 Ananya collected 150 labelled handwritten letters for a handwriting reader and keeps one fifth of them for testing. How many are left for training?
- 23.📋 Planning an AI project: which comes first — train the model or test on new photos? (a) train the model (b) test on new photos
- 24.⚖️ Two bird-song app models were tested on the same 50 bird songs. Model A got 49 right and Model B got 39 right. How many percentage points more accurate is Model A?
- 25.❌ Diya’s bird-song app was tested on 20 new bird songs and got 16 right. How many did it get wrong?
- 26.⚖️ Two fruit sorter models were tested on the same 25 fruit photos. Model A got 22 right and Model B got 21 right. How many percentage points more accurate is Model A?
- 27.⚖️ Two fruit sorter models were tested on the same 20 fruit photos. Model A got 20 right and Model B got 19 right. How many percentage points more accurate is Model A?
- 28.🎯 Rohan’s leaf identifier was tested on 25 new leaf photos and got 18 right. What is its accuracy in percent?
- 29.🎯 Meera’s bird-song app was tested on 4 new bird songs and got 2 right. What is its accuracy in percent?
- 30.🧪 You can plan a simple AI project with… (a) a secret password (b) only a big robot (c) nothing at all (d) paper and examples
- 31.📦 Rohan collected 110 labelled handwritten letters for a handwriting reader and keeps one fifth of them for testing. How many are left for training?
- 32.⚖️ Two handwriting reader models were tested on the same 20 handwritten letters. Model A got 14 right and Model B got 13 right. How many percentage points more accurate is Model A?
- 33.📋 Planning an AI project: which comes first — work out the accuracy or fix it and test again? (a) work out the accuracy (b) fix it and test again
- 34.🎯 Nisha’s bird-song app was tested on 25 new bird songs and got 13 right. What is its accuracy in percent?
- 35.📋 Planning an AI project: which comes first — recheck every label or let it learn patterns? (a) recheck every label (b) let it learn patterns
- 36.📋 Planning an AI project: which comes first — let it learn patterns or collect fruit photos? (a) let it learn patterns (b) collect fruit photos
- 37.📋 Planning an AI project: which comes first — label each example or fix it and test again? (a) label each example (b) fix it and test again
- 38.📦 Nisha collected 50 labelled waste photos for a recycling sorter and keeps one fifth of them for testing. How many are left for training?
- 39.🎯 Vihaan’s bird-song app is 60% accurate. If it is tested on 20 new bird songs, how many should it get right?
- 40.📦 Diya collected 40 labelled bird songs for a bird-song app and keeps one fifth of them for testing. How many are left for training?
Answer key
- train the model
- correct ÷ total × 100
- split train and test
- 48
- test on new photos
- correct bad labels
- 7
- work out the accuracy
- 20
- 60
- correct bad labels
- recheck every label
- 2
- tag each photo by type
- to see if it learned
- 24
- 24
- work out the accuracy
- 4
- more labelled examples
- 60
- 120
- train the model
- 20
- 4
- 4
- 5
- 72
- 50
- paper and examples
- 88
- 5
- work out the accuracy
- 52
- recheck every label
- collect fruit photos
- label each example
- 40
- 12
- 32
Free ai for students lessons, practice and worksheets at talentjr.in/ai-for-students