/en-tutor-prep-s2-ict-ai-machine-learning-intro-practice-1/
/en-tutor-prep-s2-ict-ai-machine-learning-intro-practice-1/
General Inquiry: info@TutorZone.com.hk
Office Hours: Monday to Saturday - 9:00 AM to 6:00 PM
WhatsApp Monday to Sunday - 24 hours
Hotline: +852 6828 1809
Instant Support: +852 9061 3106

S2 ICT AI practice 1: labels, accuracy and data splitting, ten items

S2 ICT AI practice 1: labels, accuracy and data splitting, ten items

  • 2026-11-20

S2 ICT AI practice 1: labels, accuracy and data splitting, ten items

Ten items start from how machine learning sits inside artificial intelligence, then cover label ratios, accuracy, train and test splits, precision, recall and steps per epoch.

S2 ICT AI practice 1: ten items S2 ICT AI practice 1: ten basic items

01 Paper details (check answers afterwards)

Item Detail
Questions 10
Suggested time 25
Difficulty mix Foundation 4 · Intermediate 2 · Challenge 2 · Error redo 2
Chapter Chapter: S2 ICT: Introduction to AI and machine learning
Level Topics cover labelled data, accuracy and data splitting.

02 Questions (10 items)

  1. FoundationHow are machine learning and artificial intelligence related in scope?
  2. FoundationOf 10,000 photos, 8,000 already carry a human label. What percentage is labelled?
  3. FoundationA model answers 170 of 200 test items correctly. What is its accuracy?
  4. Foundation9,000 records are split 8 to 2 into training and test sets. How many records each?
  5. IntermediateA confusion matrix shows 40 true positives and 10 false positives. What is the precision?
  6. IntermediateWith 40 true positives and 5 false negatives, what is the recall?
  7. ChallengeA training epoch covers 16,000 samples with batch size 32. How many steps in 20 epochs?
  8. ChallengeTo reach 90 per cent recall on 50 positive cases, how many must the model find?
  9. Error redoError: an accuracy of 95 per cent always means a good model. What is wrong?
  10. Error redoError: the test set may be reused to tune parameters until the score looks good. What is correct?

03 Answers and one-line notes

# Answer One-line note
1 Machine learning is a branch of artificial intelligence. Artificial intelligence is the broad field, machine learning is one way of training models from data, and deep learning sits inside machine learning.
2 80%. 8,000 divided by 10,000 is 0.8, which is 80 per cent.
3 85%. 170 divided by 200 equals 0.85, so the accuracy is 85 per cent.
4 7,200 for training and 1,800 for testing. 9,000 times 0.8 gives 7,200 and times 0.2 gives 1,800, and the two parts add back to the whole.
5 80%. Precision divides true positives by everything predicted positive, so 40 over 50.
6 About 88.9%. Recall is 40 over 45, which is 0.8889 or about 88.9 per cent.
7 500 steps per epoch, so 10,000 steps in total. 16,000 divided by 32 gives 500 steps per epoch, and 20 epochs give 10,000 steps.
8 45 cases. 50 multiplied by 0.9 is 45, which gives exactly 90 per cent recall.
9 Not always; with imbalanced classes precision and recall must be checked. If 95 per cent of cases share one class, always predicting it already scores 95 per cent while the rare class is never found.
10 Use a validation set for tuning and touch the test set only once at the end. Tuning against the test set feeds the answers back to the model, so the reported score becomes over-optimistic.

04 What to do after marking

  • After marking, redo only the wrong items and write down which step went wrong (formula, unit, or misread number).
  • Repeat the same item two days later; move to the next paper only when both attempts are correct.
  • Parents only need the answer and note columns; there is no need to explain every item.

Want more items on this chapter? The chapter page lists the other three papers in this set.Tutor Registration →

Questions and answers are written originally by the TutorZone team and are not taken from any school, publisher or examination authority paper. Every value is checked by computer.

Join as a tutor or coach — academic, music, sports, arts, STEM. Free to register. See how it works →