SOLUTIONS & INSTRUCTOR MANUAL
s s s
Artificial Intelligence: A Modern Approach, 4th Edition
s s s s s s
by Peter Norvig and Stuart Russell, Chapters 1 – 28
s s s s s s s s s s
,Artificial Intelligences
1 Introduction ...
ssss s s s
2 Intelligent Agents ...
ssss s s s s
II Problem-solving
s
3 Solving Problems by Searching ...
ssss s s s s s s
4 Search in Complex Environments ...
ssss s s s s s s
5 Adversarial Search and Games ...
ssss s s s s s s
6 Constraint Satisfaction Problems ...
ssss s s s s s
III Knowledge, reasoning, and planning
s s s s
7 Logical Agents ...
ssss s s s s
8 First-Order Logic ...
ssss s s s s
9 Inference in First-Order Logic ...
ssss s s s s s
10 Knowledge Representation ...
ssss s s s s
11 Automated Planning ...
ssss s s s s
IV Uncertain knowledge and reasoning
s s s s
12 Quantifying Uncertainty ...
ssss s s s s
13 Probabilistic Reasoning ...
ssss s s s s
14 Probabilistic Reasoning over Time ...
ssss s s s s s s
15 Probabilistic Programming ...
ssss s s s s
16 Making Simple Decisions ...
ssss s s s s s
17 Making Complex Decisions ...
ssss s s s s s
18 Multiagent Decision Making ...
ssss s s s s s
V Machine Learning
s s
, 19 Learning from Examples ...
ssss s s s s s
20 Learning Probabilistic Models ...
ssss s s s s s
21 Deep Learning ...
ssss s s s s
22 Reinforcement Learning ...
ssss s s s s
VI Communicating, perceiving, and acting
s s s s
23 Natural Language Processing ...
ssss s s s s s
24 Deep Learning for Natural Language Processing ...
ssss s s s s s s s s
25 Computer Vision ...
ssss s s s s
26 Robotics ...
ssss s s s
VII Conclusions
s
27 Philosophy, Ethics, and Safety of AI ...
ssss s s s s s s s s
28 The Future of AI
ssss s s s s
, EXERCISES s s
1
INTRODUCTION
Note that for many of the questions in this chapter, we give references where answers can be found
s s s s s s s s s s s s s s s s s
rather than writing them out—the full answers would be far too long.
s s s s s s s s s s s s
1.1 What Is AI?
s s s
Exercise1.1.#DEFA s
Define in your own words: (a) intelligence, (b) artificial intelligence, (c) agent, (d) ra-
s s s s s s s s s s s s s
tionality, (e) logical reasoning.
s s s s
a. Dictionary definitions of intelligence talk about “the capacity to acquire and apply s s s s s s s s s s s
knowledge” or “the faculty of thought and reason” or “the ability to comprehend and
s s s s s s s s s s s s s s
profit from experience.” These are all reasonable answers, but if we want something
s s s s s s s s s s s s s
quantifiable we would use something like “the ability to act successfully across a wide
s s s s s s s s s s s s s s
range of objectives in complex environments.”
s s s s s s
b. We define artificial intelligence as the study and construction of agent programs that
s s s s s s s s s s s s
perform well in a given class of environments, for a given agent architecture; they do the
s s s s s s s s s s s s s s s s
right thing. An important part of that is dealing with the uncertainty of what the current
s s s s s s s s s s s s s s s s
state is, what the outcome of possible actions might be, and what is it that we really desire.
s s s s s s s s s s s s s s s s s s
c. We define an agent as an entity that takes action in response to percepts from an envi-
s s s s s s s s s s s s s s s s
ronment.
s
d. We define rationality as the property of a system which does the “right thing” given what
s s s s s s s s s s s s s s s
it knows. See Section 2.2 for a more complete discussion. The basic concept is perfect
s s s s s s s s s s s s s s s
rationality; Section ?? describes the impossibility of achieving perfect rational- ity and
s s s s s s s s s s s s
proposes an alternative definition.
s s s s
e. We define logical reasoning as the a process of deriving new sentences from old, such that
s s s s s s s s s s s s s s s
the new sentences are necessarily true ifthe old ones are true. (Notice that does not refertoany
s s s s s s s s s s s s s s s s s s s
specificsyntaxorformallanguage,butitdoesrequireawell-definednotion of truth.)
s s s s s s s s s s s s s s
Exercise1.1.#TURI s
Read Turing’s original paper on AI (Turing, 1950). In the paper, he discusses several
s s s s s s s s s s s s s
objectionstohisproposedenterpriseandhistestforintelligence. Whichobjectionsstillcarry
s s s s s s s s s s s s s s
© 2023 Pearson Education, Hoboken, NJ. All rights reserved.
s s s s s s s s
s s s
Artificial Intelligence: A Modern Approach, 4th Edition
s s s s s s
by Peter Norvig and Stuart Russell, Chapters 1 – 28
s s s s s s s s s s
,Artificial Intelligences
1 Introduction ...
ssss s s s
2 Intelligent Agents ...
ssss s s s s
II Problem-solving
s
3 Solving Problems by Searching ...
ssss s s s s s s
4 Search in Complex Environments ...
ssss s s s s s s
5 Adversarial Search and Games ...
ssss s s s s s s
6 Constraint Satisfaction Problems ...
ssss s s s s s
III Knowledge, reasoning, and planning
s s s s
7 Logical Agents ...
ssss s s s s
8 First-Order Logic ...
ssss s s s s
9 Inference in First-Order Logic ...
ssss s s s s s
10 Knowledge Representation ...
ssss s s s s
11 Automated Planning ...
ssss s s s s
IV Uncertain knowledge and reasoning
s s s s
12 Quantifying Uncertainty ...
ssss s s s s
13 Probabilistic Reasoning ...
ssss s s s s
14 Probabilistic Reasoning over Time ...
ssss s s s s s s
15 Probabilistic Programming ...
ssss s s s s
16 Making Simple Decisions ...
ssss s s s s s
17 Making Complex Decisions ...
ssss s s s s s
18 Multiagent Decision Making ...
ssss s s s s s
V Machine Learning
s s
, 19 Learning from Examples ...
ssss s s s s s
20 Learning Probabilistic Models ...
ssss s s s s s
21 Deep Learning ...
ssss s s s s
22 Reinforcement Learning ...
ssss s s s s
VI Communicating, perceiving, and acting
s s s s
23 Natural Language Processing ...
ssss s s s s s
24 Deep Learning for Natural Language Processing ...
ssss s s s s s s s s
25 Computer Vision ...
ssss s s s s
26 Robotics ...
ssss s s s
VII Conclusions
s
27 Philosophy, Ethics, and Safety of AI ...
ssss s s s s s s s s
28 The Future of AI
ssss s s s s
, EXERCISES s s
1
INTRODUCTION
Note that for many of the questions in this chapter, we give references where answers can be found
s s s s s s s s s s s s s s s s s
rather than writing them out—the full answers would be far too long.
s s s s s s s s s s s s
1.1 What Is AI?
s s s
Exercise1.1.#DEFA s
Define in your own words: (a) intelligence, (b) artificial intelligence, (c) agent, (d) ra-
s s s s s s s s s s s s s
tionality, (e) logical reasoning.
s s s s
a. Dictionary definitions of intelligence talk about “the capacity to acquire and apply s s s s s s s s s s s
knowledge” or “the faculty of thought and reason” or “the ability to comprehend and
s s s s s s s s s s s s s s
profit from experience.” These are all reasonable answers, but if we want something
s s s s s s s s s s s s s
quantifiable we would use something like “the ability to act successfully across a wide
s s s s s s s s s s s s s s
range of objectives in complex environments.”
s s s s s s
b. We define artificial intelligence as the study and construction of agent programs that
s s s s s s s s s s s s
perform well in a given class of environments, for a given agent architecture; they do the
s s s s s s s s s s s s s s s s
right thing. An important part of that is dealing with the uncertainty of what the current
s s s s s s s s s s s s s s s s
state is, what the outcome of possible actions might be, and what is it that we really desire.
s s s s s s s s s s s s s s s s s s
c. We define an agent as an entity that takes action in response to percepts from an envi-
s s s s s s s s s s s s s s s s
ronment.
s
d. We define rationality as the property of a system which does the “right thing” given what
s s s s s s s s s s s s s s s
it knows. See Section 2.2 for a more complete discussion. The basic concept is perfect
s s s s s s s s s s s s s s s
rationality; Section ?? describes the impossibility of achieving perfect rational- ity and
s s s s s s s s s s s s
proposes an alternative definition.
s s s s
e. We define logical reasoning as the a process of deriving new sentences from old, such that
s s s s s s s s s s s s s s s
the new sentences are necessarily true ifthe old ones are true. (Notice that does not refertoany
s s s s s s s s s s s s s s s s s s s
specificsyntaxorformallanguage,butitdoesrequireawell-definednotion of truth.)
s s s s s s s s s s s s s s
Exercise1.1.#TURI s
Read Turing’s original paper on AI (Turing, 1950). In the paper, he discusses several
s s s s s s s s s s s s s
objectionstohisproposedenterpriseandhistestforintelligence. Whichobjectionsstillcarry
s s s s s s s s s s s s s s
© 2023 Pearson Education, Hoboken, NJ. All rights reserved.
s s s s s s s s
