SOLUTIONS & INSTRUCTOR MANUAL
ii ii ii
PR
Artificial Intelligence: A Modern Approach, 4th Edition
ii ii ii ii ii ii
by Peter Norvig and Stuart Russell, Chapters 1 – 28
ii ii ii ii ii ii ii ii ii ii
O
FD
O
C
, PR
Artificial Intelligence ii
ii ii ii ii 1 Introduction ...
ii ii ii
ii ii ii ii 2 Intelligent Agents ...
ii ii ii ii
O
II Problem-solving
ii
3 Solving Problems by Searching ...
FD
ii ii ii ii ii ii ii ii ii ii
ii ii ii ii 4 Search in Complex Environments ...
ii ii ii ii ii ii
ii ii ii ii 5 Adversarial Search and Games ...
ii ii ii ii ii ii
O
ii ii ii ii 6 Constraint Satisfaction Problems ...
ii ii ii ii ii
III Knowledge, reasoning, and planning
C
ii ii ii ii
ii ii ii ii 7 Logical Agents ...
ii ii ii ii
ii ii ii ii 8 First-Order Logic ...
ii ii ii ii
ii ii ii ii 9 Inference in First-Order Logic ...
ii ii ii ii ii
ii ii ii ii 10 Knowledge Representation ...
ii ii ii ii
ii ii ii ii 11 Automated Planning ...
ii ii ii ii
IV Uncertain knowledge and reasoning
ii ii ii ii
ii ii ii ii 12 Quantifying Uncertainty ...
ii ii ii ii
ii ii ii ii 13 Probabilistic Reasoning ...
ii ii ii ii
ii ii ii ii 14 Probabilistic Reasoning over Time ...
ii ii ii ii ii ii
ii ii ii ii 15 Probabilistic Programming ...
ii ii ii ii
ii ii ii ii 16 Making Simple Decisions ...
ii ii ii ii ii
ii ii ii ii 17 Making Complex Decisions ...
ii ii ii ii ii
ii ii ii ii 18 Multiagent Decision Making ...
ii ii ii ii ii
V Machine Learning
ii ii
,ii ii ii ii 19 Learning from Examples ...
ii ii ii ii ii
20 Learning Probabilistic Models ...
PR
ii ii ii ii ii ii ii ii ii
ii ii ii ii 21 Deep Learning ...
ii ii ii ii
ii ii ii ii 22 Reinforcement Learning ...
ii ii ii ii
O
VI Communicating, perceiving, and acting
ii ii ii ii
23 Natural Language Processing ...
FD
ii ii ii ii ii ii ii ii ii
ii ii ii ii 24 Deep Learning for Natural Language Processing ...
ii ii ii ii ii ii ii ii
ii ii ii ii 25 Computer Vision ...
ii ii ii ii
O
ii ii ii ii 26 Robotics ...
ii ii ii
VII Conclusions
C
ii
ii ii ii ii 27 Philosophy, Ethics, and Safety of AI ...
ii ii ii ii ii ii ii ii
ii ii ii ii 28 The Future of AI
ii ii ii ii
, EXERCISES 1
PR
i i ii
INTRODUCTION
O
Note iithat iifor iimany iiof iithe iiquestions iiin iithis iichapter, iiwe iigive iireferences iiwhere iianswers
iican iibe iifound iirather iithan iiwriting iithem iiout—the iifull iianswers iiwould iibe iifar iitoo iilong.
FD
1.1 i i What Is AI?
ii ii
Exercise ii1.1.#DEFA
Define iiin iiyour iiown iiwords: i i (a) iiintelligence, ii(b) iiartificial iiintelligence, ii(c) iiagent, ii(d)
O
iira- iitionality, ii(e) iilogical iireasoning.
a. Dictionary iidefinitions iiof iiintelligence iitalk iiabout ii“the iicapacity iito iiacquire iiand
C
iiapply iiknowledge” iior ii“the iifaculty iiof iithought iiand iireason” iior ii“the iiability iito
iicomprehend iiand iiprofit iifrom iiexperience.” i i These iiare iiall iireasonable iianswers,
iibut iiif iiwe iiwant iisomething iiquantifiable iiwe iiwould iiuse iisomething iilike ii“the iiability
iito iiact iisuccessfully iiacross iia iiwide iirange iiof iiobjectives iiin iicomplex iienvironments.”
b. We iidefine iiartificial iiintelligence iias iithe iistudy iiand iiconstruction iiof iiagent
iiprograms iithat iiperform iiwell iiin iia iigiven iiclass iiof iienvironments, iifor iia iigiven
iiagent iiarchitecture; iithey iido iithe iiright iithing. i i An iiimportant iipart iiof iithat iiis
iidealing iiwith iithe iiuncertainty iiof iiwhat iithe iicurrent iistate iiis, iiwhat iithe iioutcome iiof
iipossible iiactions iimight iibe, iiand iiwhat iiis iiit iithat iiwe iireally iidesire.
c. We iidefine iian iiagent iias iian iientity iithat iitakes iiaction iiin iiresponse iito iipercepts iifrom
iian iienvi- iironment.
d. We iidefine iirationality iias iithe iiproperty iiof iia iisystem iiwhich iidoes iithe ii“right iithing”
iigiven iiwhat iiit iiknows. i i See iiSection ii2.2 iifor iia iimore iicomplete iidiscussion. i i The
iibasic iiconcept iiis iiperfect iirationality; iiSection ii?? iidescribes iithe iiimpossibility iiof
iiachieving iiperfect iirational- iiity iiand iiproposes iian iialternative iidefinition.
e. We iidefine iilogical iireasoning iias iithe iia iiprocess iiof iideriving iinew iisentences iifrom iiold,
iisuch iithat iithe iinew iisentences iiare iinecessarily iitrue iiif iithe iiold iiones iiare iitrue. ii(Notice iithat
iidoes iinot iirefer iito iiany iispecific iisyntax iior iiformal iilanguage, iibut iiit iidoes iirequire iia iiwell-
defined iinotion iiof iitruth.)
Exercise ii1.1.#TURI
Read iiTuring’s iioriginal iipaper iion iiAI ii(Turing, ii1950). i i In iithe iipaper, iihe iidiscusses
iiseveral iiobjections iito iihis iiproposed iienterprise iiand iihis iitest iifor iiintelligence. ii Which iiobjections
iistill iicarry
© ii2023 iiPearson iiEducation, iiHoboken, iiNJ. iiAll iirights
iireserved.
ii ii ii
PR
Artificial Intelligence: A Modern Approach, 4th Edition
ii ii ii ii ii ii
by Peter Norvig and Stuart Russell, Chapters 1 – 28
ii ii ii ii ii ii ii ii ii ii
O
FD
O
C
, PR
Artificial Intelligence ii
ii ii ii ii 1 Introduction ...
ii ii ii
ii ii ii ii 2 Intelligent Agents ...
ii ii ii ii
O
II Problem-solving
ii
3 Solving Problems by Searching ...
FD
ii ii ii ii ii ii ii ii ii ii
ii ii ii ii 4 Search in Complex Environments ...
ii ii ii ii ii ii
ii ii ii ii 5 Adversarial Search and Games ...
ii ii ii ii ii ii
O
ii ii ii ii 6 Constraint Satisfaction Problems ...
ii ii ii ii ii
III Knowledge, reasoning, and planning
C
ii ii ii ii
ii ii ii ii 7 Logical Agents ...
ii ii ii ii
ii ii ii ii 8 First-Order Logic ...
ii ii ii ii
ii ii ii ii 9 Inference in First-Order Logic ...
ii ii ii ii ii
ii ii ii ii 10 Knowledge Representation ...
ii ii ii ii
ii ii ii ii 11 Automated Planning ...
ii ii ii ii
IV Uncertain knowledge and reasoning
ii ii ii ii
ii ii ii ii 12 Quantifying Uncertainty ...
ii ii ii ii
ii ii ii ii 13 Probabilistic Reasoning ...
ii ii ii ii
ii ii ii ii 14 Probabilistic Reasoning over Time ...
ii ii ii ii ii ii
ii ii ii ii 15 Probabilistic Programming ...
ii ii ii ii
ii ii ii ii 16 Making Simple Decisions ...
ii ii ii ii ii
ii ii ii ii 17 Making Complex Decisions ...
ii ii ii ii ii
ii ii ii ii 18 Multiagent Decision Making ...
ii ii ii ii ii
V Machine Learning
ii ii
,ii ii ii ii 19 Learning from Examples ...
ii ii ii ii ii
20 Learning Probabilistic Models ...
PR
ii ii ii ii ii ii ii ii ii
ii ii ii ii 21 Deep Learning ...
ii ii ii ii
ii ii ii ii 22 Reinforcement Learning ...
ii ii ii ii
O
VI Communicating, perceiving, and acting
ii ii ii ii
23 Natural Language Processing ...
FD
ii ii ii ii ii ii ii ii ii
ii ii ii ii 24 Deep Learning for Natural Language Processing ...
ii ii ii ii ii ii ii ii
ii ii ii ii 25 Computer Vision ...
ii ii ii ii
O
ii ii ii ii 26 Robotics ...
ii ii ii
VII Conclusions
C
ii
ii ii ii ii 27 Philosophy, Ethics, and Safety of AI ...
ii ii ii ii ii ii ii ii
ii ii ii ii 28 The Future of AI
ii ii ii ii
, EXERCISES 1
PR
i i ii
INTRODUCTION
O
Note iithat iifor iimany iiof iithe iiquestions iiin iithis iichapter, iiwe iigive iireferences iiwhere iianswers
iican iibe iifound iirather iithan iiwriting iithem iiout—the iifull iianswers iiwould iibe iifar iitoo iilong.
FD
1.1 i i What Is AI?
ii ii
Exercise ii1.1.#DEFA
Define iiin iiyour iiown iiwords: i i (a) iiintelligence, ii(b) iiartificial iiintelligence, ii(c) iiagent, ii(d)
O
iira- iitionality, ii(e) iilogical iireasoning.
a. Dictionary iidefinitions iiof iiintelligence iitalk iiabout ii“the iicapacity iito iiacquire iiand
C
iiapply iiknowledge” iior ii“the iifaculty iiof iithought iiand iireason” iior ii“the iiability iito
iicomprehend iiand iiprofit iifrom iiexperience.” i i These iiare iiall iireasonable iianswers,
iibut iiif iiwe iiwant iisomething iiquantifiable iiwe iiwould iiuse iisomething iilike ii“the iiability
iito iiact iisuccessfully iiacross iia iiwide iirange iiof iiobjectives iiin iicomplex iienvironments.”
b. We iidefine iiartificial iiintelligence iias iithe iistudy iiand iiconstruction iiof iiagent
iiprograms iithat iiperform iiwell iiin iia iigiven iiclass iiof iienvironments, iifor iia iigiven
iiagent iiarchitecture; iithey iido iithe iiright iithing. i i An iiimportant iipart iiof iithat iiis
iidealing iiwith iithe iiuncertainty iiof iiwhat iithe iicurrent iistate iiis, iiwhat iithe iioutcome iiof
iipossible iiactions iimight iibe, iiand iiwhat iiis iiit iithat iiwe iireally iidesire.
c. We iidefine iian iiagent iias iian iientity iithat iitakes iiaction iiin iiresponse iito iipercepts iifrom
iian iienvi- iironment.
d. We iidefine iirationality iias iithe iiproperty iiof iia iisystem iiwhich iidoes iithe ii“right iithing”
iigiven iiwhat iiit iiknows. i i See iiSection ii2.2 iifor iia iimore iicomplete iidiscussion. i i The
iibasic iiconcept iiis iiperfect iirationality; iiSection ii?? iidescribes iithe iiimpossibility iiof
iiachieving iiperfect iirational- iiity iiand iiproposes iian iialternative iidefinition.
e. We iidefine iilogical iireasoning iias iithe iia iiprocess iiof iideriving iinew iisentences iifrom iiold,
iisuch iithat iithe iinew iisentences iiare iinecessarily iitrue iiif iithe iiold iiones iiare iitrue. ii(Notice iithat
iidoes iinot iirefer iito iiany iispecific iisyntax iior iiformal iilanguage, iibut iiit iidoes iirequire iia iiwell-
defined iinotion iiof iitruth.)
Exercise ii1.1.#TURI
Read iiTuring’s iioriginal iipaper iion iiAI ii(Turing, ii1950). i i In iithe iipaper, iihe iidiscusses
iiseveral iiobjections iito iihis iiproposed iienterprise iiand iihis iitest iifor iiintelligence. ii Which iiobjections
iistill iicarry
© ii2023 iiPearson iiEducation, iiHoboken, iiNJ. iiAll iirights
iireserved.
