Amazon AWS Certified AI Practitioner AIF-C01 Practice Test 2
Question 10
A library has millions of articles without topic labels and wants to find naturally related themes. A colleague suggests: Train supervised classification using category labels that do not exist. Which finding BEST explains why the suggestion should be rejected?
- Speech synthesis produces spoken audio and does not inspect image patterns
- An action reward is not the same as a known class assigned to each training image
- Supervised classification requires example labels for the target categories
- The shortcut is safe because its terminology is familiar to the team
Explanation
Correct Answer
C:Supervised classification requires example labels for the target categories. This is the direct defect in the proposed action: Train supervised classification using category labels that do not exist.
Incorrect Answers
A:This concern belongs to another situation: A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing.
B:This finding is associated with a different problem: A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels.
D:Familiar wording is not proof that an action is appropriate. The actual defect is: Supervised classification requires example labels for the target categories.
Question 11
A library has millions of articles without topic labels and wants to find naturally related themes. After implementing Apply unsupervised clustering to group similar article representations, which observation BEST verifies the intended result?
- Documents with similar subjects are grouped and reviewers inspect representative clusters
- Evaluation episodes show improving cumulative reward while safety limits are enforced
- The team approves Apply unsupervised clustering to group similar article representations but does not run an operational test
- The team schedules a future test of whether documents with similar subjects are grouped and reviewers inspect representative clusters, but has no result
Explanation
Correct Answer
A:The observed result, Documents with similar subjects are grouped and reviewers inspect representative clusters, provides direct evidence. Clustering can organize unlabeled records without predefined category answers.
Incorrect Answers
B:This observation would be evidence for another case: A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels, not the current scenario.
C:The required observable condition is: Documents with similar subjects are grouped and reviewers inspect representative clusters.
D:Verify the actual outcome: Documents with similar subjects are grouped and reviewers inspect representative clusters.
Question 12
The initial scenario was: A library has millions of articles without topic labels and wants to find naturally related themes. The original control is already in place. Now editors subsequently label 20 reliable categories and want automatic categorization. Which ADDITIONAL action BEST addresses the new condition?
- Train a supervised classifier using the curated category labels
- Train supervised classification using category labels that do not exist
- Apply unsupervised clustering to group similar article representations
- Documents with similar subjects are grouped and reviewers inspect representative clusters
Explanation
Correct Answer
A:Editors subsequently label 20 reliable categories and want automatic categorization. The appropriate added action is: Train a supervised classifier using the curated category labels.
Incorrect Answers
B:This original shortcut remains unsuitable: Supervised classification requires example labels for the target categories.
C:This was the original control and is already implemented. Clustering can organize unlabeled records without predefined category answers.
D:This describes evidence for the original approach, not a new action: Documents with similar subjects are grouped and reviewers inspect representative clusters.
Question 13
A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels. Which option BEST meets the stated AI/ML or AWS requirement?
- Describe reinforcement learning with rewards for desired route outcomes
- Treat reward signals as a pre-labeled image classification dataset
- Classify image-based anomaly detection as computer vision
- Evaluate supervised imitation learning from the new labeled demonstrations
Explanation
Correct Answer
A:Reinforcement learning optimizes sequential actions using reward feedback. In this situation, choose: Describe reinforcement learning with rewards for desired route outcomes.
Incorrect Answers
B:An action reward is not the same as a known class assigned to each training image. The proposal, Treat reward signals as a pre-labeled image classification dataset, does not satisfy this need.
C:This can be valid for a different task: A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing. The current request is different and needs the selected approach.
D:This is appropriate only after a distinct change: The project later receives examples of correct routes for every junction. It is not the best initial response to a robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels.
Question 14
A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels. A colleague suggests: Treat reward signals as a pre-labeled image classification dataset. Which finding BEST explains why the suggestion should be rejected?
- The shortcut is safe because its terminology is familiar to the team
- Speech synthesis produces spoken audio and does not inspect image patterns
- An action reward is not the same as a known class assigned to each training image
- Online serving adds unnecessary immediacy and potentially different capacity costs
Explanation
Correct Answer
C:An action reward is not the same as a known class assigned to each training image. This is the direct defect in the proposed action: Treat reward signals as a pre-labeled image classification dataset.
Incorrect Answers
A:The actual defect is: An action reward is not the same as a known class assigned to each training image.
B:This finding is associated with a different problem: A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing.
D:This concern belongs to another situation: A bank serves thousands of account inquiries in one nightly job with no immediate user response requirement.
Question 15
A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels. After implementing Describe reinforcement learning with rewards for desired route outcomes, which observation BEST verifies the intended result?
- The team approves Describe reinforcement learning with rewards for desired route outcomes but does not run an operational test
- The team schedules a future test of whether evaluation episodes show improving cumulative reward while safety limits are enforced, but has no result
- Evaluation episodes show improving cumulative reward while safety limits are enforced
- The model highlights image regions with suspected abnormalities for a clinician
Explanation
Correct Answer
C:The observed result, Evaluation episodes show improving cumulative reward while safety limits are enforced, provides direct evidence. Reinforcement learning optimizes sequential actions using reward feedback.
Incorrect Answers
A:The required observable condition is: Evaluation episodes show improving cumulative reward while safety limits are enforced.
B:Verify the actual outcome: Evaluation episodes show improving cumulative reward while safety limits are enforced.
D:This observation would be evidence for another case: A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing, not the current scenario.
Question 16
The initial scenario was: A robot must learn route choices through successful and unsuccessful trial episodes rather than supplied answer labels. The original control is already in place. Now the project later receives examples of correct routes for every junction. Which ADDITIONAL action BEST addresses the new condition?
- Evaluation episodes show improving cumulative reward while safety limits are enforced
- Treat reward signals as a pre-labeled image classification dataset
- Describe reinforcement learning with rewards for desired route outcomes
- Evaluate supervised imitation learning from the new labeled demonstrations
Explanation
Correct Answer
D:The project later receives examples of correct routes for every junction. The appropriate added action is: Evaluate supervised imitation learning from the new labeled demonstrations.
Incorrect Answers
A:This describes evidence for the original approach, not a new action: Evaluation episodes show improving cumulative reward while safety limits are enforced.
B:This original shortcut remains unsuitable: An action reward is not the same as a known class assigned to each training image.
C:This was the original control and is already implemented. Reinforcement learning optimizes sequential actions using reward feedback.
Question 17
A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing. Which option BEST meets the stated AI/ML or AWS requirement?
- Describe the pixel inspection task as speech synthesis
- Choose batch inference for scheduled processing of large records
- Add speech-to-text processing for recorded dictation
- Classify image-based anomaly detection as computer vision
Explanation
Correct Answer
D:Computer vision addresses visual content such as images and image regions. In this situation, choose: Classify image-based anomaly detection as computer vision.
Incorrect Answers
A:Speech synthesis produces spoken audio and does not inspect image patterns. The proposal, Describe the pixel inspection task as speech synthesis, does not satisfy this need.
B:This can be valid for a different task: A bank serves thousands of account inquiries in one nightly job with no immediate user response requirement. The current request is different and needs the selected approach.
C:This is appropriate only after a distinct change: Clinicians then need transcripts of dictated examination notes. It is not the best initial response to a hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing.
Question 18
A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing. A colleague suggests: Describe the pixel inspection task as speech synthesis. Which finding BEST explains why the suggestion should be rejected?
- The shortcut is safe because its terminology is familiar to the team
- Batch processing cannot satisfy the live conversation response requirement
- Speech synthesis produces spoken audio and does not inspect image patterns
- Online serving adds unnecessary immediacy and potentially different capacity costs
Explanation
Correct Answer
C:Speech synthesis produces spoken audio and does not inspect image patterns. This is the direct defect in the proposed action: Describe the pixel inspection task as speech synthesis.
Incorrect Answers
A:Familiar wording is not proof that an action is appropriate. The actual defect is: Speech synthesis produces spoken audio and does not inspect image patterns.
B:This concern belongs to another situation: A mobile assistant must answer a user within seconds instead of processing an overnight queue.
D:This finding is associated with a different problem: A bank serves thousands of account inquiries in one nightly job with no immediate user response requirement.
Question 19
A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing. After implementing Classify image-based anomaly detection as computer vision, which observation BEST verifies the intended result?
- The team schedules a future test of whether the model highlights image regions with suspected abnormalities for a clinician, but has no result
- The team approves Classify image-based anomaly detection as computer vision but does not run an operational test
- The model highlights image regions with suspected abnormalities for a clinician
- The nightly results are available before the next business reporting window
Explanation
Correct Answer
C:The observed result, The model highlights image regions with suspected abnormalities for a clinician, provides direct evidence. Computer vision addresses visual content such as images and image regions.
Incorrect Answers
A:Verify the actual outcome: The model highlights image regions with suspected abnormalities for a clinician.
B:The required observable condition is: The model highlights image regions with suspected abnormalities for a clinician.
D:This observation would be evidence for another case: A bank serves thousands of account inquiries in one nightly job with no immediate user response requirement, not the current scenario.
Question 20
The initial scenario was: A hospital asks whether analyzing radiology images for anomalies is computer vision or natural language processing. The original control is already in place. Now clinicians then need transcripts of dictated examination notes. Which ADDITIONAL action BEST addresses the new condition?
- The model highlights image regions with suspected abnormalities for a clinician
- Describe the pixel inspection task as speech synthesis
- Add speech-to-text processing for recorded dictation
- Classify image-based anomaly detection as computer vision
Explanation
Correct Answer
C:Clinicians then need transcripts of dictated examination notes. The appropriate added action is: Add speech-to-text processing for recorded dictation.
Incorrect Answers
A:This describes evidence for the original approach, not a new action: The model highlights image regions with suspected abnormalities for a clinician.
B:This original shortcut remains unsuitable: Speech synthesis produces spoken audio and does not inspect image patterns.
D:This was the original control and is already implemented. Computer vision addresses visual content such as images and image regions.