Unit 2: Nature of Data and Machine Learning - Practice Quiz

1 What is data?

2 Which of the following is an example of data?

3 Which type of data includes written words?

4 Which type of data is usually stored as photographs or drawings?

5 Why is data important in artificial intelligence?

6 How can a machine learn from data?

7 What does a machine use to improve its predictions?

8 What is a label in a learning dataset?

9 If a machine studies many pictures of cats and dogs, what might it learn?

10 What happens when a machine uses learned patterns on new data?

11 What is Machine Learning?

12 What is a main goal of Machine Learning?

13 Which statement about Machine Learning is correct?

14 Does Machine Learning always require a person to write every individual rule?

15 What is the purpose of a spam filter?

16 How can a spam filter learn to recognize spam?

17 What do shopping recommendation systems suggest?

18 Which information may help a shopping system make recommendations?

19 Which statement best describes Artificial Intelligence (AI)?

20 What is automation?

21 A school records students' attendance, assignment scores, and preferred learning resources. Which statement best describes these records as data?

22 A fitness app stores step counts, exercise times, and heart-rate readings. Which feature makes these records useful for an AI system?

23 An online store has many customer records, but some ages are missing and several prices are entered incorrectly. What is the main concern?

24 Which example is best classified as unstructured data?

25 A system is shown many labeled photographs of cats and dogs before classifying a new photograph. What is the system mainly learning?

26 A model performs well on training examples but poorly on new examples. What does this most likely indicate?

27 A company wants an AI system to identify defective products. Which training approach would be most suitable if inspectors can label previous products as defective or acceptable?

28 A translation system improves after receiving corrections from users. What role do the corrections play?

29 Why should a machine-learning model be tested with data that was not used during training?

30 Which description best explains machine learning?

31 A weather application predicts rain using past weather records and current observations. Why is this an example of machine learning?

32 Which situation is least likely to require machine learning?

33 A machine-learning model gives incorrect results for groups that were poorly represented in its training data. What is the most likely cause?

34 A spam filter notices that messages containing certain patterns are often reported as unwanted. How can it use this information?

35 A spam filter marks many legitimate newsletters as spam because it learned from a biased set of examples. What would most likely help?

36 An online store recommends hiking shoes after a customer repeatedly views camping equipment. What information likely influenced the recommendation?

37 A recommendation system suggests expensive products because users often buy them after viewing similar items. What is a possible limitation of this approach?

38 Which statement best describes the relationship among artificial intelligence, machine learning, and automation?

39 A factory conveyor belt places every package into a box at a fixed time interval, regardless of package type. What is this primarily an example of?

40 A help-desk system routes messages by recognizing their topics and improving after staff correct its decisions. How should this system be classified?

41 A hospital records patients' symptoms, treatments, recovery outcomes, and the time each observation was made. Which interpretation best explains why the time information is part of the data rather than merely an administrative detail?

42 Two datasets contain the same customer transactions. Dataset A stores only product IDs and prices, while Dataset B also stores timestamps, locations, and return status. What is the strongest conclusion?

43 A company uses customer reviews to identify product defects, but most reviews are written by customers who had unusually positive or negative experiences. What property of the data is most concerning?

44 A dataset contains accurate measurements, but many records use different names for the same city. Which problem is most directly created by this inconsistency?

45 A model performs extremely well on historical training examples but poorly on new cases from the same general task. Which explanation is most plausible?

46 A spam filter is trained mostly on messages from one language and later receives messages written in another language. Why might its performance decline even if the new messages follow similar spam strategies?

47 A prediction system repeatedly makes errors for a subgroup that was rare in its training records. Which intervention most directly addresses the underlying learning problem?

48 A model predicts whether a loan will be repaid using an applicant's postal code. Even if postal code appears predictive in historical data, what is the main concern?

49 A machine-learning system is updated using user corrections, but malicious users deliberately submit false corrections. What risk does this create?

50 Which situation is the clearest example of machine learning rather than a conventional fixed-rule program?

51 Why does calling a system "machine learning" not prove that its decisions are accurate or fair?

52 A system identifies objects in photographs using many labeled examples. Which statement best describes what the system has learned?

53 A school uses a model trained on past admissions decisions to recommend future applicants. Why should the recommendations not be treated as neutral evidence of merit?

54 A spam filter begins blocking legitimate newsletters after spammers start copying newsletter-style wording. What does this example primarily demonstrate?

55 A new spam filter has very high accuracy because nearly all test messages are spam. Why could this evaluation be misleading?

56 A shopping platform recommends expensive cameras to a user who only browsed them once while researching a school project. Which limitation of recommendation learning is most evident?

57 A recommendation system mainly shows users items similar to what they already purchased. What trade-off is most likely?

58 A factory conveyor stops whenever a sensor detects excessive heat according to a rule written by an engineer. How should this system be classified most precisely?

59 Which statement best distinguishes machine learning from automation?

60 A customer-service system routes messages using fixed keywords, summarizes some messages using a learned model, and automatically sends routine confirmations. Which conclusion is most accurate?