,
, CS 7641 Final Exam Practice | Machine
Learning Practice Questions & Answers |
Georgia Tech ML Review | 2026/2027 update.
SECTION 1: SUPERVISED LEARNING (Questions 1–
15)
1. What are the two kinds of bias to worry about when designing a
classifier?
A) Approximation bias and estimation bias
B) Restriction bias and preference bias
C) Sampling bias and measurement bias
D) Inductive bias and deductive bias
Correct Answer: B
Rationale: Restriction bias refers to the set of functions the classifier can
represent (we only consider candidate functions that can be represented by
our classifier). Preference bias refers to which hypotheses the algorithm
prefers over others .
Source: CS7641 Final Exam Study Set
2. What is the preference bias of the ID3 decision tree algorithm?
A) Prefers trees with good splits at the top (greedy)
B) Prefers deeper trees
C) Prefers balanced trees
D) Prefers trees with more features
Correct Answer: A
Rationale: ID3 uses a greedy approach, preferring good splits at the top of
the tree. It also prefers shallower trees when multiple trees fit equally well .
, CS 7641 Final Exam Practice | Machine
Learning Practice Questions & Answers |
Georgia Tech ML Review | 2026/2027 update.
SECTION 1: SUPERVISED LEARNING (Questions 1–
15)
1. What are the two kinds of bias to worry about when designing a
classifier?
A) Approximation bias and estimation bias
B) Restriction bias and preference bias
C) Sampling bias and measurement bias
D) Inductive bias and deductive bias
Correct Answer: B
Rationale: Restriction bias refers to the set of functions the classifier can
represent (we only consider candidate functions that can be represented by
our classifier). Preference bias refers to which hypotheses the algorithm
prefers over others .
Source: CS7641 Final Exam Study Set
2. What is the preference bias of the ID3 decision tree algorithm?
A) Prefers trees with good splits at the top (greedy)
B) Prefers deeper trees
C) Prefers balanced trees
D) Prefers trees with more features
Correct Answer: A
Rationale: ID3 uses a greedy approach, preferring good splits at the top of
the tree. It also prefers shallower trees when multiple trees fit equally well .
