Scribd
Upload
10 views8 pages

Unit 4 Aisnotesdududu

Uploaded by

thakku2005
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
10 views8 pages

Unit 4 Aisnotesdududu

Uploaded by

thakku2005
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

UNIT

Expert System Avchite


h a ti s an exprt Systerm)
(omputer Prog us ed to So|ve
problems
proude deuston Comp lex
ma king
abhuy ike a man
expert
It loes his by
KB uSirg th extrachg
veasoni knowledee
inference from
ng inerence l
acording fo e
queries
&fart o AI

ESA

|. Knwlede bare

Stores donaunspeeifre kvowled


as swlls, ats heurishes
Deveped usi ng exte nsive ntera chan
w hnan experts updoted os
IS vcu lable
S Cx Hedi cal di cgnosis Synptoms,
diseases possible heatmetS.
ns
2. UscY hterae.
’AllowS eneKpert ugers to
i nterad w system, posing queries
uiievi g explaunah ons
’ Intertae si mpifi es eomplex
in
eries A presets ests
Understte format
3. Inyevente Engine
’ Brain or ES
ules to kB ho
derive (onsions
Two types:
forward cheining
4Whet will happen ny)
facts to concusion
x Backard Chaining
why'd it happen)
gal back to knowh
7acts
4. Explounation System
’ hecsoning & jushficti ons foY
System's decisi onS, ch enhane,
trust in 34stem s otputs.
’(xplains why a paricla 0ln
Was
NLs appied.

S. Knasledse Aguesihon mo del


- updotes haintouns kR

Relihes eRishng know ledge to


Ohsw stem erai ng. urrent
dtteeive.
Rele base d Avchiteture
Commn tyre
tnowtede epreserted aS
’ Ingeren
enonne processes
s e s to arrive ot Con wsrong
A Simple to impleunt
’ Can
become complex
as no of teg
0ncreacS

Common y used in YcIN for ne diceuy


diagnosis
2 Frane- bared Architectue
knouw ledoe represented
StruchUe caed FRAHES
Similar too obie ts in odp.
oDP
tach praune represents a concept
W atribtes and vaues
J This ar chútetwe is erriien
for domains hat neuire
hierarcni cal ergan3on
3. Black bonrd orchite cture
S Mliple independent sub-systems
Work 2yeher b sharing
Comghon workspece coled
4 Blackbonrd".
3 Subsystems contri buh part1ol
sclns
collarahon
to BB wich are reHned hru
BB
> TAeal lor complen, multi disiplirarg
olorms
4. Hodel based Aechi le ctu
Uss model of system or do nin i
bper at es in.
Simules behewior
by comparihg otuol per formune w
nodel hthawi

Nost commony wred in


Xpe
troube shooh
teA
5. Hy brid architeete
Com bihe scveral types of architectm.
to leverage strernghs oF each
’ AUovws fo
power fw problem sowing apprcoch,
pati uloyin eomplex domaing
howleche Kiyr eenhhon
Dtemnes Erert Systemns
wknowled
stored 4
accesed yucturn
strucured,
Meno ds
) Rale based
Yepresenation

->Represenhirg Knowlsde as

’Usuly
"iY - en
in the form of
-’ This steemonts
arprocch is
rechve or
domai ns whre KnoWLe
be de can
Cleay arhCuated as ulg.
2) Semnanhc Networks
Knouw ledye greph of nodes
no
(conepts) e e
blw con upts).
dges (re lahonshp
<S Alo wscomp telatronshp
hiearhes
3) frames
-’ Jal structUes wsed to
eprese sterot ped itucufi ong,
o bje s oy cnhties
5 Each framne = slots (attibutes)

ns are uuseu oy
fYanos
Ovgan3 re krewledye in
structusal ierarthicat dorans.
’ Alows experts o deal w/
wner anuty ^preision.
’ hepresertirg concpts wch
aen t black -and- white ht
exis + in deyrea

5) houction wles
-S Know ledheset of prod. Jus
fhat speç fy achi ons bac d On
londi ons
’ Usd in systems hnat egnie
procedwl Knw hcr, bke
houle Shooing or dagroshies

6) Logic based reyrej entation


’redi cate prqosi honal Asi
used to tomaize complex
eleonSi ps A make infer
fowerel
’ Bud computa oney
expensive, equirig sophsh cated
alyoris for easoning
Expevt System Shel
a S¥ware de velop nent enironment
ProVdes asic Come ononts hecesse ry
4or bulding
deelopus to code orm Sir
afch
) TnYeren
Ergine
COmes w pre bullt 16
hat can py Aocad. ules o
herishcs to kß to yerive oncuSi ons

)2 kB stiucture
3Sheu rovides a anework or
au deve lapers to (npu conaunspecae
Knaw ledge oten in forms oF ues,
fames on ooic statemants
DAllous dyelopers to focus on cap turing
Cxpevt kno Lade wo woryns cbot
lmpemonting undylying oLgovihs
3) User Ihteryace
tS shel inudes tDls to ma ke UT,
h to interact
enalinog noh- cpest
sqstem eagiy
SuT Jupporis queryht
osut s disp lay
eXpluQhi ong
Explou nahon oles,
m
4)
explain ngy how if arrived at
a par ieula conusion.
)V trust understandi

s)S) Kitow ledge aqç ss han toots

inpt hmeir knaw Ledge dire ct


me system
y his impiies froces_ ok
wpdahing and mauntan kR

Exaumple
CLIPS
JESS

5 hreaty reduces dev. hme


aser mci ntenene,

Scalathuty

You might also like