1
AWS Certified AI
Practitioner (AIF-C01) —
Advanced Practice Exam
150 Multiple-Choice
Questions with Answers
and Rationales a well
detailed one 2025 /
2026 written and graded
A+ upgraded
, 2
Domain 1: Fundamentals of AI and ML (20%) — Questions 1–30
Question 1
A financial institution is building a system to identify fraudulent credit card transactions in real
time. The system must flag suspicious transactions with high precision to minimize false alarms
that would inconvenience customers. Which machine learning approach is most appropriate for
this use case?
A. Unsupervised clustering to group transactions by amount
B. Supervised classification using labeled historical fraud data
C. Reinforcement learning to optimize fraud detection policies
D. Regression analysis to predict transaction amounts
-” detailed answer 100 % correct :-”B
Rationale: Fraud detection is a binary classification problem where the model learns from
labeled historical data (fraudulent vs. legitimate transactions). Classification models can be
tuned for high precision by adjusting the decision threshold. Unsupervised clustering (A) lacks
labeled guidance; reinforcement learning (C) is for sequential decision-making; regression (D)
predicts continuous values, not categories.
Question 2
A data scientist observes that a machine learning model achieves 99% accuracy on training data
but only 78% accuracy on a held-out test set. What is the most likely explanation for this
performance gap?
A. The model is underfitting the training data
B. The test set contains outliers that should be removed
C. The model is overfitting the training data
D. The training data has insufficient features
, 3
-” detailed answer 100 % correct :-”C
Rationale: Overfitting occurs when a model learns noise and patterns specific to the training
data rather than generalizable patterns, resulting in high training accuracy but poor
performance on unseen data. Underfitting (A) would show poor performance on both sets.
Outliers (B) and insufficient features (D) would typically affect both training and test
performance.
Question 3
A retail company wants to segment its customer base into distinct groups based on purchasing
behavior without any predefined categories. Which machine learning technique should be
used?
A. Classification
B. Regression
C. Clustering
D. Reinforcement learning
-” detailed answer 100 % correct :-”C
Rationale: Clustering is an unsupervised learning technique used to discover natural groupings
in unlabeled data. Since the company has no predefined categories, unsupervised clustering is
appropriate. Classification (A) requires labeled data; regression (B) predicts continuous values;
reinforcement learning (D) uses reward-based learning.
Question 4
Which statement correctly differentiates between supervised learning, unsupervised learning,
and reinforcement learning?
A. Supervised learning uses unlabeled data; unsupervised learning uses labeled data
B. Reinforcement learning learns from a reward signal; supervised learning uses labeled input-
output pairs
C. Unsupervised learning requires a reward function; reinforcement learning finds patterns in
data
D. All three learning types require labeled training data
-” detailed answer 100 % correct :-”B
, 4
Rationale: Reinforcement learning involves an agent learning optimal actions through rewards
and penalties. Supervised learning uses labeled input-output pairs. Unsupervised learning finds
patterns in unlabeled data. Options A and C reverse these definitions; D is incorrect because
unsupervised learning does not require labels.
Question 5
A hospital has developed a machine learning model to predict patient readmission risk within
30 days of discharge. The model uses features including age, diagnosis, number of previous
admissions, and medication adherence. What type of machine learning problem is this?
A. Regression
B. Classification
C. Clustering
D. Dimensionality reduction
-” detailed answer 100 % correct :-”B
Rationale: Predicting a categorical outcome (readmitted or not readmitted within 30 days) is a
classification problem. Regression (A) predicts continuous values; clustering (C) finds natural
groupings; dimensionality reduction (D) reduces feature space.
Question 6
Which of the following best describes the relationship between artificial intelligence (AI),
machine learning (ML), and deep learning (DL)?
A. AI is a subset of ML, which is a subset of DL
B. ML is a subset of AI, and DL is a subset of ML
C. DL is a subset of AI, and ML is a separate discipline
D. AI, ML, and DL are three distinct and unrelated fields
-” detailed answer 100 % correct :-”B
Rationale: Machine learning is a subset of artificial intelligence that enables systems to learn
from data. Deep learning is a subset of machine learning that uses neural networks with
multiple layers. This hierarchical relationship is foundational to understanding AI concepts.
Question 7
AWS Certified AI
Practitioner (AIF-C01) —
Advanced Practice Exam
150 Multiple-Choice
Questions with Answers
and Rationales a well
detailed one 2025 /
2026 written and graded
A+ upgraded
, 2
Domain 1: Fundamentals of AI and ML (20%) — Questions 1–30
Question 1
A financial institution is building a system to identify fraudulent credit card transactions in real
time. The system must flag suspicious transactions with high precision to minimize false alarms
that would inconvenience customers. Which machine learning approach is most appropriate for
this use case?
A. Unsupervised clustering to group transactions by amount
B. Supervised classification using labeled historical fraud data
C. Reinforcement learning to optimize fraud detection policies
D. Regression analysis to predict transaction amounts
-” detailed answer 100 % correct :-”B
Rationale: Fraud detection is a binary classification problem where the model learns from
labeled historical data (fraudulent vs. legitimate transactions). Classification models can be
tuned for high precision by adjusting the decision threshold. Unsupervised clustering (A) lacks
labeled guidance; reinforcement learning (C) is for sequential decision-making; regression (D)
predicts continuous values, not categories.
Question 2
A data scientist observes that a machine learning model achieves 99% accuracy on training data
but only 78% accuracy on a held-out test set. What is the most likely explanation for this
performance gap?
A. The model is underfitting the training data
B. The test set contains outliers that should be removed
C. The model is overfitting the training data
D. The training data has insufficient features
, 3
-” detailed answer 100 % correct :-”C
Rationale: Overfitting occurs when a model learns noise and patterns specific to the training
data rather than generalizable patterns, resulting in high training accuracy but poor
performance on unseen data. Underfitting (A) would show poor performance on both sets.
Outliers (B) and insufficient features (D) would typically affect both training and test
performance.
Question 3
A retail company wants to segment its customer base into distinct groups based on purchasing
behavior without any predefined categories. Which machine learning technique should be
used?
A. Classification
B. Regression
C. Clustering
D. Reinforcement learning
-” detailed answer 100 % correct :-”C
Rationale: Clustering is an unsupervised learning technique used to discover natural groupings
in unlabeled data. Since the company has no predefined categories, unsupervised clustering is
appropriate. Classification (A) requires labeled data; regression (B) predicts continuous values;
reinforcement learning (D) uses reward-based learning.
Question 4
Which statement correctly differentiates between supervised learning, unsupervised learning,
and reinforcement learning?
A. Supervised learning uses unlabeled data; unsupervised learning uses labeled data
B. Reinforcement learning learns from a reward signal; supervised learning uses labeled input-
output pairs
C. Unsupervised learning requires a reward function; reinforcement learning finds patterns in
data
D. All three learning types require labeled training data
-” detailed answer 100 % correct :-”B
, 4
Rationale: Reinforcement learning involves an agent learning optimal actions through rewards
and penalties. Supervised learning uses labeled input-output pairs. Unsupervised learning finds
patterns in unlabeled data. Options A and C reverse these definitions; D is incorrect because
unsupervised learning does not require labels.
Question 5
A hospital has developed a machine learning model to predict patient readmission risk within
30 days of discharge. The model uses features including age, diagnosis, number of previous
admissions, and medication adherence. What type of machine learning problem is this?
A. Regression
B. Classification
C. Clustering
D. Dimensionality reduction
-” detailed answer 100 % correct :-”B
Rationale: Predicting a categorical outcome (readmitted or not readmitted within 30 days) is a
classification problem. Regression (A) predicts continuous values; clustering (C) finds natural
groupings; dimensionality reduction (D) reduces feature space.
Question 6
Which of the following best describes the relationship between artificial intelligence (AI),
machine learning (ML), and deep learning (DL)?
A. AI is a subset of ML, which is a subset of DL
B. ML is a subset of AI, and DL is a subset of ML
C. DL is a subset of AI, and ML is a separate discipline
D. AI, ML, and DL are three distinct and unrelated fields
-” detailed answer 100 % correct :-”B
Rationale: Machine learning is a subset of artificial intelligence that enables systems to learn
from data. Deep learning is a subset of machine learning that uses neural networks with
multiple layers. This hierarchical relationship is foundational to understanding AI concepts.
Question 7
