SOLUTIONS & INSTRUCTOR MANUAL
Artificial Intelligence: A Modern Approach, 4th Edition
by Peter Norvig and Stuart Russell, Chapters 1 – 28
,Artificial Intelligence nw
nwnwnwnw 1 Introduction ...
nw nw nw
nwnwnwnw 2 Intelligent Agents ...
nw nw nw nw
II Problem-solving
nw
nwnwnwnw 3 Solving Problems by Searching ...
nw nw nw nw nw nw
nwnwnwnw 4 Search in Complex Environments ...
nw nw nw nw nw nw
nwnwnwnw 5 Adversarial Search and Games ...
nw nw nw nw nw nw
nwnwnwnw 6 Constraint Satisfaction Problems ...
nw nw nw nw nw
III Knowledge, reasoning, and planning
nw nw nw nw
nwnwnwnw 7 Logical Agents ...
nw nw nw nw
nwnwnwnw 8 First-Order Logic ...
nw nw nw nw
nwnwnwnw 9 Inference in First-Order Logic ...
nw nw nw nw nw
nwnwnwnw 10 Knowledge Representation ...
nw nw nw nw
nwnwnwnw 11 Automated Planning ...
nw nw nw nw
IV Uncertain knowledge and reasoning
nw nw nw nw
nwnwnwnw 12 Quantifying Uncertainty ...
nw nw nw nw
nwnwnwnw 13 Probabilistic Reasoning ...
nw nw nw nw
nwnwnwnw 14 Probabilistic Reasoning over Time ...
nw nw nw nw nw nw
nwnwnwnw 15 Probabilistic Programming ...
nw nw nw nw
nwnwnwnw 16 Making Simple Decisions ...
nw nw nw nw nw
nwnwnwnw 17 Making Complex Decisions ...
nw nw nw nw nw
nwnwnwnw 18 Multiagent Decision Making ...
nw nw nw nw nw
V Machine Learning
nw nw
, 19 Learning from Examples ...
nwnwnwnw nw nw nw nw nw
20 Learning Probabilistic Models ...
nwnwnwnw nw nw nw nw nw
21 Deep Learning ...
nwnwnwnw nw nw nw nw
22 Reinforcement Learning ...
nwnwnwnw nw nw nw nw
VI Communicating, perceiving, and acting
nw nw nw nw
23 Natural Language Processing ...
nwnwnwnw nw nw nw nw nw
24 Deep Learning for Natural Language Processing ...
nwnwnwnw nw nw nw nw nw nw nw nw
25 Computer Vision ...
nwnwnwnw nw nw nw nw
26 Robotics ...
nwnwnwnw nw nw nw
VII Conclusions
nw
27 Philosophy, Ethics, and Safety of AI ...
nwnwnwnw nw nw nw nw nw nw nw nw
28 The Future of AI
nwnwnwnw nw nw nw nw
, EXERCISES n w n w
1
INTRODUCTION
Note that for many of the questions in this chapter, we give references where answers can
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
be found rather than writing them out—the full answers would be far too long.
nw nw nw nw nw nw nw nw nw nw nw nw nw
1.1 What Is AI?
n w nw nw
Exercise 1.1.#DEFA nw
Define in your own words: (a) intelligence, (b) artificial intelligence, (c) agent, (d) ra-
nw nw nw nw n w nw nw nw nw nw nw nw nw
tionality, (e) logical reasoning.
nw nw nw nw
a. Dictionary definitions of intelligence talk about “the capacity to acquire and apply
nw nw nw nw nw nw nw nw nw nw nw nw
knowledge” or “the faculty of thought and reason” or “the ability to comprehend a
nw nw nw nw nw nw nw nw nw nw nw nw nw
nd profit from experience.” These are all reasonable answers, but if we want some
nw nw nw n w nw nw nw nw nw nw nw nw nw
thing quantifiable we would use something like “the ability to act successfully across
nw nw nw nw nw nw nw nw nw nw nw nw
a wide range of objectives in complex environments.”
nw nw nw nw nw nw nw nw
b. We define artificial intelligence as the study and construction of agent programs t
nw nw nw nw nw nw nw nw nw nw nw nw
hat perform well in a given class of environments, for a given agent architecture; t
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
hey do the right thing. An important part of that is dealing with the uncertainty of
nw nw nw nw n w nw nw nw nw nw nw nw nw nw nw n
what the current state is, what the outcome of possible actions might be, and what
w nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
is it that we really desire.
nw nw nw nw nw
c. We define an agent as an entity that takes action in response to percepts from an en
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
vi- ronment. nw
d. We define rationality as the property of a system which does the “right thing” giv
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
en what it knows. See Section 2.2 for a more complete discussion. The basic con
nw nw nw n w nw nw nw nw nw nw nw n w nw nw
cept is perfect rationality; Section ?? describes the impossibility of achieving perfect
nw nw nw nw nw nw nw nw nw nw nw nw
rational- ity and proposes an alternative definition.nw nw nw nw nw nw
e. We define logical reasoning as the a process of deriving new sentences from old, suc
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
h that the new sentences are necessarily true if the old ones are true. (Notice that does
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
not refer to any specific syntax or formal language, but it does require a well-
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
defined notion of truth.) nw nw nw
Exercise 1.1.#TURI nw
Read Turing’s original paper on AI (Turing, 1950). In the paper, he discusses several o
nw nw nw nw nw nw nw n w nw nw nw nw nw nw
bjections to his proposed enterprise and his test for intelligence. Which objections still carry
nw nw nw nw nw nw nw nw nw nw nw nw nw
© 2023 Pearson Education, Hoboken, NJ. All rights reserv
nw nw nw nw nw nw nw nw
ed.
Artificial Intelligence: A Modern Approach, 4th Edition
by Peter Norvig and Stuart Russell, Chapters 1 – 28
,Artificial Intelligence nw
nwnwnwnw 1 Introduction ...
nw nw nw
nwnwnwnw 2 Intelligent Agents ...
nw nw nw nw
II Problem-solving
nw
nwnwnwnw 3 Solving Problems by Searching ...
nw nw nw nw nw nw
nwnwnwnw 4 Search in Complex Environments ...
nw nw nw nw nw nw
nwnwnwnw 5 Adversarial Search and Games ...
nw nw nw nw nw nw
nwnwnwnw 6 Constraint Satisfaction Problems ...
nw nw nw nw nw
III Knowledge, reasoning, and planning
nw nw nw nw
nwnwnwnw 7 Logical Agents ...
nw nw nw nw
nwnwnwnw 8 First-Order Logic ...
nw nw nw nw
nwnwnwnw 9 Inference in First-Order Logic ...
nw nw nw nw nw
nwnwnwnw 10 Knowledge Representation ...
nw nw nw nw
nwnwnwnw 11 Automated Planning ...
nw nw nw nw
IV Uncertain knowledge and reasoning
nw nw nw nw
nwnwnwnw 12 Quantifying Uncertainty ...
nw nw nw nw
nwnwnwnw 13 Probabilistic Reasoning ...
nw nw nw nw
nwnwnwnw 14 Probabilistic Reasoning over Time ...
nw nw nw nw nw nw
nwnwnwnw 15 Probabilistic Programming ...
nw nw nw nw
nwnwnwnw 16 Making Simple Decisions ...
nw nw nw nw nw
nwnwnwnw 17 Making Complex Decisions ...
nw nw nw nw nw
nwnwnwnw 18 Multiagent Decision Making ...
nw nw nw nw nw
V Machine Learning
nw nw
, 19 Learning from Examples ...
nwnwnwnw nw nw nw nw nw
20 Learning Probabilistic Models ...
nwnwnwnw nw nw nw nw nw
21 Deep Learning ...
nwnwnwnw nw nw nw nw
22 Reinforcement Learning ...
nwnwnwnw nw nw nw nw
VI Communicating, perceiving, and acting
nw nw nw nw
23 Natural Language Processing ...
nwnwnwnw nw nw nw nw nw
24 Deep Learning for Natural Language Processing ...
nwnwnwnw nw nw nw nw nw nw nw nw
25 Computer Vision ...
nwnwnwnw nw nw nw nw
26 Robotics ...
nwnwnwnw nw nw nw
VII Conclusions
nw
27 Philosophy, Ethics, and Safety of AI ...
nwnwnwnw nw nw nw nw nw nw nw nw
28 The Future of AI
nwnwnwnw nw nw nw nw
, EXERCISES n w n w
1
INTRODUCTION
Note that for many of the questions in this chapter, we give references where answers can
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
be found rather than writing them out—the full answers would be far too long.
nw nw nw nw nw nw nw nw nw nw nw nw nw
1.1 What Is AI?
n w nw nw
Exercise 1.1.#DEFA nw
Define in your own words: (a) intelligence, (b) artificial intelligence, (c) agent, (d) ra-
nw nw nw nw n w nw nw nw nw nw nw nw nw
tionality, (e) logical reasoning.
nw nw nw nw
a. Dictionary definitions of intelligence talk about “the capacity to acquire and apply
nw nw nw nw nw nw nw nw nw nw nw nw
knowledge” or “the faculty of thought and reason” or “the ability to comprehend a
nw nw nw nw nw nw nw nw nw nw nw nw nw
nd profit from experience.” These are all reasonable answers, but if we want some
nw nw nw n w nw nw nw nw nw nw nw nw nw
thing quantifiable we would use something like “the ability to act successfully across
nw nw nw nw nw nw nw nw nw nw nw nw
a wide range of objectives in complex environments.”
nw nw nw nw nw nw nw nw
b. We define artificial intelligence as the study and construction of agent programs t
nw nw nw nw nw nw nw nw nw nw nw nw
hat perform well in a given class of environments, for a given agent architecture; t
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
hey do the right thing. An important part of that is dealing with the uncertainty of
nw nw nw nw n w nw nw nw nw nw nw nw nw nw nw n
what the current state is, what the outcome of possible actions might be, and what
w nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
is it that we really desire.
nw nw nw nw nw
c. We define an agent as an entity that takes action in response to percepts from an en
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
vi- ronment. nw
d. We define rationality as the property of a system which does the “right thing” giv
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
en what it knows. See Section 2.2 for a more complete discussion. The basic con
nw nw nw n w nw nw nw nw nw nw nw n w nw nw
cept is perfect rationality; Section ?? describes the impossibility of achieving perfect
nw nw nw nw nw nw nw nw nw nw nw nw
rational- ity and proposes an alternative definition.nw nw nw nw nw nw
e. We define logical reasoning as the a process of deriving new sentences from old, suc
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
h that the new sentences are necessarily true if the old ones are true. (Notice that does
nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw nw
not refer to any specific syntax or formal language, but it does require a well-
nw nw nw nw nw nw nw nw nw nw nw nw nw nw
defined notion of truth.) nw nw nw
Exercise 1.1.#TURI nw
Read Turing’s original paper on AI (Turing, 1950). In the paper, he discusses several o
nw nw nw nw nw nw nw n w nw nw nw nw nw nw
bjections to his proposed enterprise and his test for intelligence. Which objections still carry
nw nw nw nw nw nw nw nw nw nw nw nw nw
© 2023 Pearson Education, Hoboken, NJ. All rights reserv
nw nw nw nw nw nw nw nw
ed.
