Artificial Intelligence: A Modern Approach, 4th Edition
By Peter Norvig And Stuart Russell, Chapters 1 – 28
,Artificial Intelligence g
gggg 1 Introduction ...
g g g
gggg 2 Intelligent Agents ...
g g g g
II Problem-solving
g
gggg 3 Solving Problems by Searching ...
g g g g g g
gggg 4 Search in Complex Environments ...
g g g g g g
gggg 5 Adversarial Search and Games ...
g g g g g g
gggg 6 Constraint Satisfaction Problems ...
g g g g g
III Knowledge, reasoning, and planning
g g g g
gggg 7 Logical Agents ...
g g g g
gggg 8 First-Order Logic ...
g g g g
gggg 9 Inference in First-Order Logic ...
g g g g g
gggg 10 Knowledge Representation ...
g g g g
gggg 11 Automated Planning ...
g g g g
IV Uncertain knowledge and reasoning
g g g g
gggg 12 Quantifying Uncertainty ...
g g g g
gggg 13 Probabilistic Reasoning ...
g g g g
gggg 14 Probabilistic Reasoning over Time ...
g g g g g g
gggg 15 Probabilistic Programming ...
g g g g
gggg 16 Making Simple Decisions ...
g g g g g
gggg 17 Making Complex Decisions ...
g g g g g
gggg 18 Multiagent Decision Making ...
g g g g g
V Machine Learning
g g
,gggg 19 Learning from Examples ...
g g g g g
gggg 20 Learning Probabilistic Models ...
g g g g g
gggg 21 Deep Learning ...
g g g g
gggg 22 Reinforcement Learning ...
g g g g
VI Communicating, perceiving, and acting
g g g g
gggg 23 Natural Language Processing ...
g g g g g
gggg 24 Deep Learning for Natural Language Processing ...
g g g g g g g g
gggg 25 Computer Vision ...
g g g g
gggg 26 Robotics ...
g g g
VII Conclusions
g
gggg 27 Philosophy, Ethics, and Safety of AI ...
g g g g g g g g
gggg 28 The Future of AI
g g g g
, EXERCISES g g
1
INTRODUCTION
Notegthatgforgmanygofgthegquestionsgingthisgchapter,gweggivegreferencesgwhereganswersgc
angbegfoundgrathergthangwritinggthemgout—thegfullganswersgwouldgbegfargtooglong.
1.1 What Is AI?
g g g
Exerciseg1.1.#DEFA
Definegingyourgowngwords:g (a)gintelligence,g(b)gartificialgintelligence,g(c)gagent,g(d)gra-
gtionality, g(e)glogicalgreasoning.
a. Dictionarygdefinitionsgofgintelligencegtalkgaboutg“thegcapacitygtogacquiregandgapp
lygknowledge”gorg“thegfacultygofgthoughtgandgreason”gorg“thegabilitygtogcompreh
endgandgprofitgfromgexperience.”g Thesegaregallgreasonableganswers,gbutgifgwegwa
ntgsomethinggquantifiablegwegwouldgusegsomethingglikeg“thegabilitygtogactgsuccessf
ullygacrossgagwidegrangegofgobjectivesgingcomplexgenvironments.”
b. Wegdefinegartificialgintelligencegasgthegstudygandgconstructiongofgagentgprogramsg
thatgperformgwellgingaggivengclassgofgenvironments,gforgaggivengagentgarchitecture
;gtheygdogthegrightgthing.g Angimportantgpartgofgthatgisgdealinggwithgtheguncertaint
ygofgwhatgthegcurrentgstategis,gwhatgthegoutcomegofgpossiblegactionsgmightgbe,gan
dgwhatgisgitgthatgwegreallygdesire.
c. Wegdefinegangagentgasgangentitygthatgtakesgactiongingresponsegtogperceptsgfromgang
envi-gronment.
d. Wegdefinegrationalitygasgthegpropertygofgagsystemgwhichgdoesgtheg“rightgthing”ggi
vengwhatgitgknows.g SeegSectiong2.2gforgagmoregcompletegdiscussion.g Thegbasicgco
nceptgisgperfectgrationality;gSectiong??gdescribesgthegimpossibilitygofgachievinggperf
ectgrational-gitygandgproposesgangalternativegdefinition.
e. Wegdefineglogicalgreasoninggasgthegagprocessgofgderivinggnewgsentencesgfromgold,gs
uchgthatgthegnewgsentencesgaregnecessarilygtruegifgthegoldgonesgaregtrue.g(Noticegtha
tgdoesgnotgrefergtoganygspecificgsyntaxgorgformalglanguage,gbutgitgdoesgrequiregagwell
-definedgnotiongofgtruth.)
Exerciseg1.1.#TURI
ReadgTuring’sgoriginalgpapergongAIg(Turing,g1950).g Ingthegpaper,ghegdiscussesgseveralgobjec
tionsgtoghisgproposedgenterprisegandghisgtestgforgintelligence.gWhichgobjectionsgstillgcarry
©g2023gPearsongEducation,gHoboken,gNJ.gAllgrightsgreserved
.