Introduction

What is Artificial Intelligence ?

• John McCarthy, who coined (invented) the term Artificial
Intelligence in 1956,defines it as "the science and engineering of
making intelligent machines", especially intelligent computer
programs.
• Artificial Intelligence (AI) is the intelligence of machines and
the branch
of computer science that aims to create it.
• Intelligence is the computational part of the ability to achieve
goals
in the world. Varying kinds and degrees of intelligence occur in
people, many animals and some machines.
AI isthestudy of the mentalfaculties (sense) through the
use of computational models.
 AIisthe study and design of intelligentagents, where an
intelligentagentis asystem that perceives its
environment and takes actions thatmaximize itschances
ofsuccess.
ThedefinitionsofAIoutlinedintextbooks
Intelligence
 Related to tasks involving higher mental processes
Examples:
creativity, solving problems, pattern recognition,
classification, learning, induction, deduction, optimization,
languageprocessing, knowledgeandmanymore.
Intelligence is the computational part of the ability to achieve
goals.
IntelligentBehavior
perceiving one’senvironment.
ƒ Acting in complex environments.
ƒ Learning and understanding from experience.
ƒ Reasoning to solve problemsand discover hidden
knowledge,ƒ Knowledge applying successfullyin new
situations.
ƒ Thinking abstractly, using analogies.
ƒ Communicating with others
ƒ Creativity, Ingenuity, Expressive-ness, Curiosity
UnderstandingAI
ƒ Howknowledge isacquired, represented, and stored;
ƒ Howintelligent behavior isgenerated and learned ;
ƒ How motives, emotions, and priorities are developed and
used;
ƒ Howsensory signalsare transformedintosymbols;
ƒ How symbols are manipulated to perform logic, to reason
aboutpastand plan forfuture;
ƒ How mechanisms of intelligence produce the phenomena
of illusion, belief, hope, fear, dreams, kindness and love
HardorStrongAI
 Generally, artificial intelligence research aims to create AI
thatcanreplicate human intelligence completely .
 Strong AI refers to a machine that approaches or
supersedeshumanintelligence,
 Ifitcandotypicallyhumantasks,
 Ifitcanapplyawiderangeofbackgroundknowledgeand
 Ifithassomedegree ofself-consciousness.
 Strong AI aims to build machines whose overall
intellectual ability is indistinguishable from that of a
human being.
SoftorWeakAI
 Weak AI refers to the use of software to study or
accomplish specific problem solving or reasoning tasks
that do not encompass (cover) the full range of human
cognitive abilities.
Weak AI does not achieve self-awareness; it
demonstrates wide range of human-level cognitive
abilities; it is merely an intelligent, a specific problem-
solver
CognitiveScience
 Aims to develop, explore and evaluate theories of how the
mind worksthrough the use ofcomputationalmodels.
 The important is not what is done but how it is done;
means intelligent behavior is not enough, the program
must
operate in an intelligent manner.
GoalsofAI
 The definitionsof AI givesfour possiblegoalsto pursue :
Systems that thinklikehumans.
 Systems that thinkrationally.
Systems that act like humans
Systems that act rationally
GeneralAIGoal
ƒ Replicatehumanintelligence : stilladistantgoal.

ƒ Solveknowledgeintensivetasks.
ƒMakeanintelligentconnection betweenperceptionandaction.
ƒ Enhance human-human, human-computer and computer to computer
interaction / communication.
 EngineeringbasedAIGoal
ƒ Developconcepts, theoryandpracticeofbuildingintelligentmachines

Emphasisisonsystembuilding.
SciencebasedAIGoal
Develop concepts, mechanisms and vocabulary to understand biological
intelligentbehavior.
Emphasisisonunderstandingintelligentbehavior.
Cognitivescience : Thinkhuman-like
An excitingnewefforttomake
computersthink; thatit is, the machines
withminds, inthe fullandliteral sense.
Focusisnot just onbehavior andI/O, but
looksat reasoning process.
 Computational modelastohowresults
were obtained.
Goalisnot just to produce human-like
behaviorbutto produce a sequence of
stepsofthe reasoningprocess, similar to
the stepsfollowed bya human in solving
the same task.
Laws ofThought : ThinkRationally
 Thestudy ofmentalfacultiesthroughtheuseofcomputational
models; thatitis, thestudy ofthe computationsthatmakeit
possibletoperceive, reason, and act.
 Focusison inferencemechanismsthatareprovablycorrectand
guaranteean optimalsolution.
 Developsystemsofrepresentationto allowinference
(conclusion) tobelike “Socratesisaman. Allmen aremortal
(human, earthly) . Therefore Socratesismortal.”
 Goalisto formalizethe reasoning processasa systemoflogical
rulesandproceduresforinference.
TuringTest : ActHuman-like
 The artof creating machinesthatperform functions
requiring intelligence when performed by people; thatit
isthe study of, how to make computers do thingswhich at
themomentpeopledo better.
 Focusison action, and notintelligentbehaviorcentered
around representation of theworld.
 A Behavioristapproach, isnot concerned with how to get
resultsbutto the similarity to whathuman results are
Example : TuringTest
3 rooms contain: aperson, acomputer, and an interrogator.
 Theinterrogatorcan communicatewith theother 2 byteletype (to avoid
themachineimitatetheappearanceorvoiceoftheperson).
 Theinterrogatortries to determinewhich is thepersonand which is
themachine.
Themachinetries tofooltheinterrogatorto believethatitis the
human, and theperson alsotries to convincetheinterrogatorthatit
is thehuman.
 Ifthemachinesucceeds in foolingtheinterrogator, then conclude
thatthemachineis intelligent.

Goalis todevelop systems thatarehuman-like.

RationalAgent : ActRationally
 Triesto explain and emulate intelligentbehavior in terms
of
computational processes; that itisconcerned with the
automation ofintelligence.
 Focusison systemsthatactsufficiently if notoptimally in
all situations;
 Goal is to develop systemsthat are rational and
sufficient.
AITechniques
Various techniquesthathave evolved , can be applied toa variety
of AI tasks.
The techniques are concerned with how we represent , manipulate
and reason with knowledge in order to solve problems .

Examples
• Techniques, not all "intelligent " butused to behave asintelligent
ƒ 1. Describe and match ƒ 2. Goalreduction
ƒ 3. Constraint satisfaction ƒ 4. Tree Searching
ƒ 5. Generate and test ƒ [Link] based systems
Biology-inspiredAItechniquesare
currentlypopular
ƒ Artificial Neural Networksƒ
Genetic Algorithms
AntColony Optimization Algorithm
ParticleSwarm Optimization
Cuckoo Search Algorithm
DescribeandMatch
 Model is adescription of a system’s behavior .
 Finite state model consists of a set of states , a set of input events and
the relations between them. Given a current state and an input event
you can determinethe next current stateof themodel .
 Computation model is a finite state machine . It includes of a set of
states, a set of start states, an input alphabet, and a transition
function which maps input symbols and current states to anext state .
 Representation of computational system include start and end state
descriptions and a set of possible transition rules that might be applied .
Problem is to find theappropriatetransition rules .
 Transition relation: If a pair of states (S , S ') is such that one move
takes the system from S to S', then the transition relation is represented
by S => S
State-transition system is called deterministic if every state has at
most one successor; it is called non-deterministic if at least one state
has morethan one successor.
Puzzle : TowersofHanoiwithonly 2 disks
Examplesof some possibletransitionsbetween statesare
shown for theTowers ofHanoi puzzle.
Example
GoalReduction
 Goal-reduction procedures are a special case of the procedural
representationsofknowledgein AI; logic-based representations.
 The process involves the hierarchical sub-division of goals into sub-goals,
until the sub-goals which have an immediate solution are reached
andsaid “goalhasbeensatisfied”.
 Goal-reduction process is illustrated in the form of AND/OR tree drawn
upside-down.
◊ Goal levels : Higher-level goals are higher in the tree, and lower level goals
arelowerin thetree.
◊ Arcs are directed from a higher-to-lower level node represents the
reductionofhigher-levelgoalto lower-levelsub-goal.
◊ Nodesatthebottomofthetreerepresentirreducibleaction goals.
 An AND-OR tree/graph structure can represent relations between goals
and sub-goals, alternative sub-goals and conjoint sub-goals.
ExampleGoal Reduction
AND-OR tree/graph structure to representfactssuch as
“enjoyment”,
“earning/save money”, “old age” etc.
Theabove AND-ORtree/graphstructure describes
◊ Hierarchical relationships between goals and sub-goals
The“goingon strike”isa sub-goalof “earningmore money”, isa
sub-goal of“improving standardof living”, is a sub-goal of
“improving enjoyment of life”.
◊ Alternative ways of trying to solvea goal
The“goingon strike”and“increasingproductivity” are alternative
ways of tryingto“earn more money” (increase pay).
e.g.: “improving standardofliving” and“working less hard” are
alternative ways oftryingto “improve enjoyment oflife”.
◊ Conjoint sub-goals
To “provide for old age”, not onlyneed to “earnmore money”, but
aswell need to “save money”.
ConstraintSatisfactionTechniques
Constraintisa logicalrelation among variables. e.g.
“circle isinside thesquare” – Theconstraintsrelate
objectswithoutprecisely specifying their positions;
moving any one, the relation isstillmaintained.
 Constraintsatisfaction isa processof finding asolution to
a setofconstraints– theconstraintsexpressallowed
valuesfor variablesand finding solution isevaluation of
thesevariablesthat satisfiesall constraints.
ConstraintSatisfactionProblem (CSP)
anditssolution
◊ AConstraint Satisfaction Problem (CSP) consistsof :
o Variables, afiniteset X = {x1 , . . . , xn } ,
o Domain, afiniteset Di of possible valueswhicheachvariablexi
cantake,
o Constraints, aset of valuesthat thevariablescansimultaneously
satisfytheconstraint (e.g. D1 != D2)
◊ Asolutionto aCSPis an assignment ofavalue fromits domainto
every variable satisfying every constraint ; that couldbe :
o onesolution, withnopreferenceastowhichone,
o all solutions,
o anoptimal, or agoodsolution - Constraint OptimizationProblem
(COP).
N-Queens puzzle
Problem : Given any integer N, place N queens on N*N
chessboard satisfying constraint that no two queens
threaten each other. (a queen threatens other queens on
samerow, column and diagonal).
Solution :
AcademicDisciplines relevant toAI
Logic, methodsofreasoning, mind asphysical Philosophy 
.system, foundationsoflearning, language, rationality

Formalrepresentation and proof, algorithms, Mathematics 

computation, (un)decidability, (in)tractability

modeling uncertainty, learning fromdataProbability/Statistics 

utility, decision theory, rational economicagents Economics 

.neuronsasinformationprocessing units Neuroscience 

howdo peoplebehave, perceive, processcognitivePsychology/ 

.information, representknowledgeCognitiveScience

building fastcomputers Computer 

engineering

design systemsthat maximizean objective Control theory 

function overtime

knowledgerepresentation, grammars Linguistics 

ABriefHistoryof AI
Thegestation (growth) ofAI (1943 - 1956):
- 1943: McCulloch&Pitts: Booleancircuitmodelofbrain.
- 1950: Turing’s“ComputingMachineryandIntelligence”.
- 1956: McCarthy’sname“ArtificialIntelligence” adopted.
Earlyenthusiasm, greatexpectations (1952 - 1969):
The early years of AI were full of successes—in a limited way
- EarlysuccessfulAIprograms: Samuel’s checkers ( likeagame)
- Newell and Simon's early success was followed up with the General Problem
Solver orGPS.
GPS was probably the first program to embody the "thinking
humanly" approach
Herbert Gelernter (1959) constructed the Geometry Theorem Prover, which was
able toprove theoremsthat manystudentsofmathematicswouldfindquite tricky
ABriefHistoryofAI
Adoseofreality (1966 - 1974):
- AIdiscoveredcomputational complexity.
- Neural networkresearchalmostdisappeared
after
Minsky&Papert’sbookin 1969.
Knowledge-basedsystems (1969 - 1979):
- 1969: DENDRALbyBuchananetal..
- 1976: MYCINbyShortliffle.
- 1979: PROSPECTORbyDudaetal..
ABriefHistoryofAI
AIbecomesan industry (1980 - 1988):
- Expert systemsindustry booms.
- 1981: Japan’s 10-yearFifth Generation project.

Thereturn ofNNsand novel AI (1986 - present):

- Mid 80’s: Back-propagation learningalgorithm reinvented.
- Expert systemsindustry busts.
- 1988: Resurgence ofprobability.
- 1988: Novel AI (ALife, GAs, Soft Computing, …).
- 1995: Agentseverywhere.
- 2003: Human-level AIbackon theagenda.
ABriefHistoryof AI
AI adopts thescientific method (1987 –present)
Recent years have seen a revolution in both the content and the methodology of
workinartificialintelligence.
It is now more common to build on existing theories than to propose brand -new
ones, to base claims on rigorous theorems or hard experimental evidence rather
than on intuition(perception) , and to show relevance to real-world
applications rather than toy examples.
In terms of methodology, AI has finally come firmly under the scientific
method.
To be accepted, hypotheses must be subjected to rigorous empirical
experiments, and the results must
beanalyzedstatistically for their importance (Cohen , 1995).
It is now possible to replicate experiments by using shared repositories of test
data andcode.