CPE 126 INTRODUCTION TO ARTIFICIAL
INTELLIGENCE MODULE 2 EXAM | QUESTIONS AND
ANSWERS | 2026 UPDATE - MAPÚA UNIVERSITY.
146 Questions with Answers and Detailed Rationales
100 PERCENT GUARANTEED PASS
INSTANT DOWNLOAD ANSWERS INCLUDED
IMPORTANCE OF THIS DOCUMENT
This comprehensive examination preparation guide has been meticulously developed to help you succeed in the
CPE 126 INTRODUCTION TO ARTIFICIAL INTELLIGENCE MODULE 2 EXAM | QUESTIONS AND ANSWERS
| 2026 UPDATE - MAPÚA UNIVERSITY.. It contains 146 carefully selected questions that reflect the most current
exam content and testing strategies. Each question is accompanied by a correct answer and a detailed rationale
that explains the underlying pathophysiology, pharmacology, or clinical reasoning.
Self-Assessment – Test your knowledge and Exam Preparation – Familiarize yourself with the
identify areas requiring further question format and content
study areas
Concept Reinforcement – Deepen your Confidence Building – Develop test-taking
understanding through strategies and reduce
evidence-based exam anxiety
rationales
Time Management – Practice answering
questions under simulated
exam conditions
Review Summary 146 Questions
Foundations - Application - CPE 126 Introduction TO Artificial Intelligence Module 2 AND 2026 Update -
MAP A University Introduction TO Artificial Intelligence CPE 126 Module 2 Search Knowledge Representation
AND Machine Learning Foundations Undergraduate YEAR 2-3 Sophomore/junior MAP A University CPE
Department
All answers with rationales
,Table of Contents
Content Area Questions Key Topics
Introduction TO Artificial 1-25 Student, Accuracy, Trains, Training, Directly
Intelligence
Intelligent Agents 26-50 Search, CPE 126 S, Pruning, Network, Training
Problem Solving BY 51-75 Student, Training, Accuracy, Network, Dataset
Searching
Uninformed Search 76-100 Search, Change, Module, Directly, Value
Strategies
Informed Search Strategies 101-125 Search, Agent, Inference, Learning, Logic
Local Search AND 126-146 Student, Network, Search, Problem, Inference
Optimization
TOTAL 146 All questions include answers and detailed rationales
,Section A - Introduction TO Artificial Intelligence
Q1.
A Mapúa robotics team models a warehouse robot's navigation as a state-space graph
where every action has strictly positive cost. Which search strategy is guaranteed to
return a cost-optimal path while remaining complete on finite graphs?
A. Depth-First Search (DFS) B. Breadth-First Search (BFS) regardless of
edge weights
C. Uniform-Cost Search (UCS) D. Greedy Best-First Search
Correct: C - Uniform-Cost Search (UCS)
Rationale:UCS expands the lowest-cost frontier node, so it is complete and optimal for
non-negative step costs. BFS is optimal only when all step costs are equal, DFS is neither
optimal nor complete on infinite graphs, and Greedy Best-First Search uses only a heuristic
and can return suboptimal paths.
Why the other answers are wrong:
A. DFS can dive down a deep, expensive branch and is not guaranteed to find the cheapest
path.
B. BFS is optimal only if every edge has the same cost, which is not guaranteed here.
D. Greedy Best-First ignores path cost and can be misled by an optimistic heuristic.
Reference: Russell, S. & Norvig, P. (2021). Artificial Intelligence: A Modern Approach, 4th Ed., Ch. 3.4
Q2.
In a two-player zero-sum game tree, alpha-beta pruning is applied with a perfect
move-ordering heuristic. What is the best-case reduction in the number of nodes
evaluated compared to plain minimax for a tree of branching factor b and depth d?
A. From O(b^d) to O(b^(d/2)) B. From O(b^d) to O(d^b)
C. From O(b^d) to O(b-d) D. No asymptotic reduction; only a
constant-factor speedup
Correct: A - From O(b^d) to O(b^(d/2))
Rationale:With optimal move ordering, alpha-beta pruning reduces the effective branching
factor to its square root, giving O(b^(d/2)) - equivalent to doubling the search depth for the
same cost. The other options misstate the asymptotic behavior of alpha-beta pruning.
Why the other answers are wrong:
B. O(d^b) inverts the roles of b and d and is not the alpha-beta bound.
C. O(b-d) describes linear-time graph traversal, not game-tree pruning.
D. Alpha-beta does achieve an asymptotic reduction, not merely a constant speedup.
Page 3
, Section A - Introduction TO Artificial Intelligence
Reference: Russell, S. & Norvig, P. (2021). AIMA, 4th Ed., Ch. 5.3
Q3.
A knowledge base contains: (P -> Q), (Q -> R), and P. Using forward chaining, which atom
is derived first, and what is the final entailed conclusion?
A. R is derived first; Q is the final conclusion B. Q is derived first; R is the final conclusion
C. P is re-derived; Q and R are never D. Both Q and R are derived simultaneously
entailed in a single step
Correct: B - Q is derived first; R is the final conclusion
Rationale:Forward chaining fires rules whose premises are satisfied: P triggers (P !’ Q),
yielding Q; then Q triggers (Q -> R), yielding R. Thus Q is derived first and R is the final
conclusion. The other options misorder the inference steps or deny entailment.
Why the other answers are wrong:
A. R cannot be derived before Q, since (Q -> R) requires Q as a premise.
C. P is already a fact; forward chaining does proceed to derive Q and R.
D. Forward chaining is sequential here; R depends on Q being established first.
Reference: Russell, S. & Norvig, P. (2021). AIMA, 4th Ed., Ch. 7.5 (Forward Chaining)
Q4.
A Mapúa student trains a decision tree on 500 labeled soil-moisture readings but achieves
99% training accuracy and 62% test accuracy. Which remedy most directly addresses this
gap?
A. Increase tree depth further to capture B. Apply pruning or limit maximum depth,
more patterns and gather more training data
C. Switch to a linear regression model with D. Remove cross-validation and evaluate
no regularization only on the training set
Correct: B - Apply pruning or limit maximum depth, and gather more training data
Rationale:The large train-test gap indicates overfitting (high variance). Pruning or bounding
tree depth reduces variance, and additional data improves generalization. The other choices
either worsen overfitting or mask the problem.
Why the other answers are wrong:
A. Deeper trees increase variance and worsen the overfitting already observed.
C. An unregularized linear model may underfit or overfit and does not directly fix the tree's
variance.
D. Removing validation hides the generalization gap instead of resolving it.
Reference: Géron, A. (2022). Hands-On Machine Learning, 3rd Ed., Ch. 6 (Decision Trees)
Page 4
INTELLIGENCE MODULE 2 EXAM | QUESTIONS AND
ANSWERS | 2026 UPDATE - MAPÚA UNIVERSITY.
146 Questions with Answers and Detailed Rationales
100 PERCENT GUARANTEED PASS
INSTANT DOWNLOAD ANSWERS INCLUDED
IMPORTANCE OF THIS DOCUMENT
This comprehensive examination preparation guide has been meticulously developed to help you succeed in the
CPE 126 INTRODUCTION TO ARTIFICIAL INTELLIGENCE MODULE 2 EXAM | QUESTIONS AND ANSWERS
| 2026 UPDATE - MAPÚA UNIVERSITY.. It contains 146 carefully selected questions that reflect the most current
exam content and testing strategies. Each question is accompanied by a correct answer and a detailed rationale
that explains the underlying pathophysiology, pharmacology, or clinical reasoning.
Self-Assessment – Test your knowledge and Exam Preparation – Familiarize yourself with the
identify areas requiring further question format and content
study areas
Concept Reinforcement – Deepen your Confidence Building – Develop test-taking
understanding through strategies and reduce
evidence-based exam anxiety
rationales
Time Management – Practice answering
questions under simulated
exam conditions
Review Summary 146 Questions
Foundations - Application - CPE 126 Introduction TO Artificial Intelligence Module 2 AND 2026 Update -
MAP A University Introduction TO Artificial Intelligence CPE 126 Module 2 Search Knowledge Representation
AND Machine Learning Foundations Undergraduate YEAR 2-3 Sophomore/junior MAP A University CPE
Department
All answers with rationales
,Table of Contents
Content Area Questions Key Topics
Introduction TO Artificial 1-25 Student, Accuracy, Trains, Training, Directly
Intelligence
Intelligent Agents 26-50 Search, CPE 126 S, Pruning, Network, Training
Problem Solving BY 51-75 Student, Training, Accuracy, Network, Dataset
Searching
Uninformed Search 76-100 Search, Change, Module, Directly, Value
Strategies
Informed Search Strategies 101-125 Search, Agent, Inference, Learning, Logic
Local Search AND 126-146 Student, Network, Search, Problem, Inference
Optimization
TOTAL 146 All questions include answers and detailed rationales
,Section A - Introduction TO Artificial Intelligence
Q1.
A Mapúa robotics team models a warehouse robot's navigation as a state-space graph
where every action has strictly positive cost. Which search strategy is guaranteed to
return a cost-optimal path while remaining complete on finite graphs?
A. Depth-First Search (DFS) B. Breadth-First Search (BFS) regardless of
edge weights
C. Uniform-Cost Search (UCS) D. Greedy Best-First Search
Correct: C - Uniform-Cost Search (UCS)
Rationale:UCS expands the lowest-cost frontier node, so it is complete and optimal for
non-negative step costs. BFS is optimal only when all step costs are equal, DFS is neither
optimal nor complete on infinite graphs, and Greedy Best-First Search uses only a heuristic
and can return suboptimal paths.
Why the other answers are wrong:
A. DFS can dive down a deep, expensive branch and is not guaranteed to find the cheapest
path.
B. BFS is optimal only if every edge has the same cost, which is not guaranteed here.
D. Greedy Best-First ignores path cost and can be misled by an optimistic heuristic.
Reference: Russell, S. & Norvig, P. (2021). Artificial Intelligence: A Modern Approach, 4th Ed., Ch. 3.4
Q2.
In a two-player zero-sum game tree, alpha-beta pruning is applied with a perfect
move-ordering heuristic. What is the best-case reduction in the number of nodes
evaluated compared to plain minimax for a tree of branching factor b and depth d?
A. From O(b^d) to O(b^(d/2)) B. From O(b^d) to O(d^b)
C. From O(b^d) to O(b-d) D. No asymptotic reduction; only a
constant-factor speedup
Correct: A - From O(b^d) to O(b^(d/2))
Rationale:With optimal move ordering, alpha-beta pruning reduces the effective branching
factor to its square root, giving O(b^(d/2)) - equivalent to doubling the search depth for the
same cost. The other options misstate the asymptotic behavior of alpha-beta pruning.
Why the other answers are wrong:
B. O(d^b) inverts the roles of b and d and is not the alpha-beta bound.
C. O(b-d) describes linear-time graph traversal, not game-tree pruning.
D. Alpha-beta does achieve an asymptotic reduction, not merely a constant speedup.
Page 3
, Section A - Introduction TO Artificial Intelligence
Reference: Russell, S. & Norvig, P. (2021). AIMA, 4th Ed., Ch. 5.3
Q3.
A knowledge base contains: (P -> Q), (Q -> R), and P. Using forward chaining, which atom
is derived first, and what is the final entailed conclusion?
A. R is derived first; Q is the final conclusion B. Q is derived first; R is the final conclusion
C. P is re-derived; Q and R are never D. Both Q and R are derived simultaneously
entailed in a single step
Correct: B - Q is derived first; R is the final conclusion
Rationale:Forward chaining fires rules whose premises are satisfied: P triggers (P !’ Q),
yielding Q; then Q triggers (Q -> R), yielding R. Thus Q is derived first and R is the final
conclusion. The other options misorder the inference steps or deny entailment.
Why the other answers are wrong:
A. R cannot be derived before Q, since (Q -> R) requires Q as a premise.
C. P is already a fact; forward chaining does proceed to derive Q and R.
D. Forward chaining is sequential here; R depends on Q being established first.
Reference: Russell, S. & Norvig, P. (2021). AIMA, 4th Ed., Ch. 7.5 (Forward Chaining)
Q4.
A Mapúa student trains a decision tree on 500 labeled soil-moisture readings but achieves
99% training accuracy and 62% test accuracy. Which remedy most directly addresses this
gap?
A. Increase tree depth further to capture B. Apply pruning or limit maximum depth,
more patterns and gather more training data
C. Switch to a linear regression model with D. Remove cross-validation and evaluate
no regularization only on the training set
Correct: B - Apply pruning or limit maximum depth, and gather more training data
Rationale:The large train-test gap indicates overfitting (high variance). Pruning or bounding
tree depth reduces variance, and additional data improves generalization. The other choices
either worsen overfitting or mask the problem.
Why the other answers are wrong:
A. Deeper trees increase variance and worsen the overfitting already observed.
C. An unregularized linear model may underfit or overfit and does not directly fix the tree's
variance.
D. Removing validation hides the generalization gap instead of resolving it.
Reference: Géron, A. (2022). Hands-On Machine Learning, 3rd Ed., Ch. 6 (Decision Trees)
Page 4
