CS 7641 SL Unit Quiz Solutions| Questions and
Answers Latest Updated 2026/2027 | Georgia
Institute of Technology
INTRODUCTION
Welcome to the CS 7641 Supervised Learning Unit Quiz study guide — your
complete, 2026/2027-updated resource for passing the SL Unit Quiz in
Georgia Tech's OMSCS Machine Learning course on the first attempt. This
guide contains practice questions and answers, each paired with a bolded
correct answer and a detailed rationale covering decision trees,
regression and classification, neural networks, instance-based learning,
computational learning theory, and ensemble methods and feature
selection. Whether you are preparing for the SL Unit Quiz, reviewing
during the course, or using this as a rapid-reference for machine
learning concepts, this guide mirrors the style, difficulty, and content of
the actual quiz.
SECTION 1: DECISION TREES (ID3, INFORMATION GAIN, ENTROPY,
PRUNING)
Q1.
What is the ID3 algorithm?
A. A greedy algorithm that selects the attribute with the highest
,information gain at each step
B. A lazy learning algorithm
C. A clustering algorithm
D. A reinforcement learning algorithm
Answer: A) A greedy algorithm that selects the attribute with the
highest information gain at each step
Rationale: ID3 builds decision trees by recursively selecting the
attribute that maximizes information gain.
Q2.
What is entropy in the context of decision trees?
A. A measure of impurity in a dataset
B. A measure of accuracy
C. A measure of speed
D. A measure of memory usage
Answer: A) A measure of impurity in a dataset
Rationale: Entropy quantifies the uncertainty or impurity in a set of
examples.
Q3.
What is the entropy of a dataset where all examples have the same label?
A. 0
B. 1
C. 0.5
D. 2
,Answer: A) 0
Rationale: A pure dataset has zero entropy.
Q4.
What is the entropy of a dataset with equal classes?
A. 1
B. 0
C. 0.5
D. 2
Answer: A) 1
Rationale: Maximum entropy for binary classification is 1 when
classes are equally distributed.
Q5.
What is information gain?
A. The reduction in entropy after splitting on an attribute
B. The increase in entropy after splitting
C. The accuracy of a decision tree
D. The depth of a decision tree
Answer: A) The reduction in entropy after splitting on an attribute
Rationale: Information gain measures how much an attribute
reduces uncertainty.
Q6.
Which attribute does ID3 choose at each step?
A. The attribute with the highest information gain
B. The attribute with the lowest information gain
, C. A random attribute
D. The first attribute in the dataset
Answer: A) The attribute with the highest information gain
Rationale: ID3 is a greedy algorithm that maximizes information
gain.
Q7.
When does a decision tree stop growing?
A. When all examples have the same label
B. When all attributes are used
C. When the tree reaches maximum depth
D. All of the above
Answer: A) When all examples have the same label
Rationale: A pure node has zero entropy and does not need to be
split further.
Q8.
What is a leaf node in a decision tree?
A. A node where all examples have the same label
B. The root node
C. An internal node
D. A branch
Answer: A) A node where all examples have the same label
Rationale: Leaf nodes represent the final classification.
Q9.
What happens when two attributes have the same highest information
Answers Latest Updated 2026/2027 | Georgia
Institute of Technology
INTRODUCTION
Welcome to the CS 7641 Supervised Learning Unit Quiz study guide — your
complete, 2026/2027-updated resource for passing the SL Unit Quiz in
Georgia Tech's OMSCS Machine Learning course on the first attempt. This
guide contains practice questions and answers, each paired with a bolded
correct answer and a detailed rationale covering decision trees,
regression and classification, neural networks, instance-based learning,
computational learning theory, and ensemble methods and feature
selection. Whether you are preparing for the SL Unit Quiz, reviewing
during the course, or using this as a rapid-reference for machine
learning concepts, this guide mirrors the style, difficulty, and content of
the actual quiz.
SECTION 1: DECISION TREES (ID3, INFORMATION GAIN, ENTROPY,
PRUNING)
Q1.
What is the ID3 algorithm?
A. A greedy algorithm that selects the attribute with the highest
,information gain at each step
B. A lazy learning algorithm
C. A clustering algorithm
D. A reinforcement learning algorithm
Answer: A) A greedy algorithm that selects the attribute with the
highest information gain at each step
Rationale: ID3 builds decision trees by recursively selecting the
attribute that maximizes information gain.
Q2.
What is entropy in the context of decision trees?
A. A measure of impurity in a dataset
B. A measure of accuracy
C. A measure of speed
D. A measure of memory usage
Answer: A) A measure of impurity in a dataset
Rationale: Entropy quantifies the uncertainty or impurity in a set of
examples.
Q3.
What is the entropy of a dataset where all examples have the same label?
A. 0
B. 1
C. 0.5
D. 2
,Answer: A) 0
Rationale: A pure dataset has zero entropy.
Q4.
What is the entropy of a dataset with equal classes?
A. 1
B. 0
C. 0.5
D. 2
Answer: A) 1
Rationale: Maximum entropy for binary classification is 1 when
classes are equally distributed.
Q5.
What is information gain?
A. The reduction in entropy after splitting on an attribute
B. The increase in entropy after splitting
C. The accuracy of a decision tree
D. The depth of a decision tree
Answer: A) The reduction in entropy after splitting on an attribute
Rationale: Information gain measures how much an attribute
reduces uncertainty.
Q6.
Which attribute does ID3 choose at each step?
A. The attribute with the highest information gain
B. The attribute with the lowest information gain
, C. A random attribute
D. The first attribute in the dataset
Answer: A) The attribute with the highest information gain
Rationale: ID3 is a greedy algorithm that maximizes information
gain.
Q7.
When does a decision tree stop growing?
A. When all examples have the same label
B. When all attributes are used
C. When the tree reaches maximum depth
D. All of the above
Answer: A) When all examples have the same label
Rationale: A pure node has zero entropy and does not need to be
split further.
Q8.
What is a leaf node in a decision tree?
A. A node where all examples have the same label
B. The root node
C. An internal node
D. A branch
Answer: A) A node where all examples have the same label
Rationale: Leaf nodes represent the final classification.
Q9.
What happens when two attributes have the same highest information
