4. DrsCm about Vauous Steps for data mintng tork .

Hoo
A%sociation mules, comiftcation Can bebe wsed for analyts ,
Student potmane
Data mining is proen o sttng thoough Loouge
sts to fdentfy atterna and dota
solve buntnes pnblema though melattonhps that can help
data analysrs.
clasiffcation
a) cuutení ng
3) Assbcí atíon rules
4) Regrnien
5) Duíatíon Joutie dete ctio
’camiatten Dt fnvolves amigning aa clas label to each
?stanee în a dat et ba ed on (ts teates. Itt
is to build model thot
qoal
acwataly prredut elas labels
new fstanes based
Dn t treatus.
is o tuoo typs bínaoy elantfatón yOnly too
close Soch as Spaun , not span'
mutti lan clasifauin
’ moe tnon

objects ?rto clasts o cinilan DbËects

Dbt. (s Knowr
cwtuing
>Asoialón Rles speclio latonips ameng the
D data fberns. Tt shoos hoo ugLenty a itemst
Dccurs fn a asaeton.
Suppont 1imAa hord Regunty
(reaslUres the Co llection
i Ju

OccUt
tathn as penceotaqe o all thansadtions
Congidence s- e tells how teguntly' an item o(lwy

opet to otht Ttemset.

Eegemfon a CUUe

tne vaodows data pofnts. and osed t predict any contou

Valucd ttibute. Used fo finanal tore casting
Outlfa/beutatíon deteatuén".- Outiey dtectien fe proces a
detectfng beetlí or data point tfat fs ban auay
avag dependng Dn what we cUe to to qt by remov Tng
them thom analyre to prewent any potentíal
sso cíatfen mles ío Studet cteuofng
ptomance i- Co uwwe
selectíen. i .e freguenty
tak yes. tidy tabit: tssoiaton
coes.

qraden nd
Getween shuy habits ç
sudy stonges 9 pomo them amonq all studen
-Atendence Gpyomance: Tdentify Lnes btw atendene
ptovmae
Can hiqhiaht innportane o_ veqular endene o acodemic Succe
Extacuoúculas Aciitles- Rweols how s tudents cngaqed in Speufc
eytiacwlar acttuitie tends to ptorm bette
cloniatien r s in Student Po7ormone i
woounna
Students
S4steen i- betldfnqa nodel, sc hool can predrct

frteveriens to hetp -thereo impove.

Ctudent Pot1ng'- Crecate poilso, Sucentl Sludents, (dntiby'ng
Common haits thot cobibute 4hoin Succen, en (ovage otbes Students
Ao asopt fmilon behavtouoy.
Couwe Recormnon dalícon Bsed on post pYrmance this model Can
recomend aitable Cosen fur ctubeals t hat align oith thein
cthenatG, (nctets.
 2

Juntty tne stotement toith emample 'Domain Cn ouoledg helps to

enthact Inooledge in a beten
betten uway fn data míning".
Name T3
Ghenden Feven
Jack
Cough T
N N. N
Mauy P N
M
N
Compu'te the d(Jack, May) dcMay,J) and d(Tacle,J m)
thom above data set.
Genden is symmtie atbubute , rematninq atthibtes ene
Cymituc bincoy
lit vas and P Ge 4 and Value N is O.
Stem
2trts (Asamdtsae)
t Stt
(Syometbác) Sn
p

(35milonity)
fordSaeK, Many)
Aegmbic
Jack
3 (e)
d(Gat Mag) o.5

1+)
 R
l(9) Ol) =0.5
Jack
lo 3 (t)
d(Tace, Sim)

lo, Doman Knoolcde helps to erbract tnouoletge in a btttey

tor enample ohen apíect afns to
Customer
pochaning bchouiows Ros Qn Onlne clothina Sbore. without
doma Khaledge he data analysts might appoach the tak
pluiy tom teethni cad pempectfve , appling enec data
mint nq lqoithms do dataset. tHousew en if the
noedae about tashien fnduty
posses
e doan
onlne retail they
etact more
stth domatn nocoledqe anayts can'r
I: Pecune selectien!- They Can choose
cottqorí s, Ctosenal trends, featiie ke clothino
rgences , fmp rovfnq areLay
tlgotthre Cutoizathen'r hey coan adapt
kon specifte patens in tanhion algoíthm to
tndy auch as

3. nterpretalin
for îytance if sud alen
Sales
fs pbsewed domaln Enoole de míqh alv
bite to cole
weathe Totfaes han fauttey aata
pornt;
o)
h.0-|h
9

()naddns

rodens (9)
gtt.0 gt> %9-tE X
oddns to(a)
) Data
elcontng'
4hat
They Can ond f ecitiy data in(onsislen
6e lniue to
Cente ual insiqht:- Dx
ndithy
aloe analyst< to add
conlexl l hudi
Tey Can deteíne new thend r
Jong
8) BoTaqçted Deciseen Matínq' ith domain inoights the bntinc
Sye (an make fnomed odeersíons uth ay inbodie(ny
pvdutt line based on (dentfred Custom Peaenee.
finally hís alloo to enchance 2uaity e veleuant entocteA
inihts hdance d undetandng
he data lcading betten de cistens stiateqfes sothin înalujtby
fnd te teguent tem sets fpom tolloofng Gansaetion dataset
wsfng aprierf algorfhm oitth mfnimu Suppost =3
Dtemset

{a,d,e 3
Sab, c,e
3 { a,brd, e?
a,c, d, eB

b, dy e?

{ a,d,e
{aib,et
6) Contuet
set
ctuong
wstng apri
asoration rules tor above eguent î ten
algoithm itfh conhrdenee H52
 Suppot Ca,e,b)
t - 0-5
tuppot (a,ey
Suppot (a,b,e) 0.16O.a5=

Suppovt (ab)
fos fa, 4, e uppotfa
{at d,e? - Sppot a,4,e)
Suppot (a)
euppot (a , e
erppot (dy
Suppot( aid, e) o.52o 5

Suppot (aid, e) =064<o.35 = G2X

Srppot (a,e)
Sad iey hppot(a,d,e)
-|o.45
Cuppot (a.dy
Most stuong asoiotion ules ahei

49) Consid, to tlouir St

anacions genecte amociaten
ues eoing Cuppot - 507.
ind Condence
Growtk
TYanactien L'st o iteme

1,,32,S4

ot

narge the tatle otte rte ro

uppot er
drgrordra
T:

13,T4
 altu thayacfen 5
tiee

3 /.

ottes a Atuón & inal

T3:5.

5 Discws the scope 4 applicatên data manfoa fa Banting

Data matng has oide Scepe oy appications in bantrng fndutey

Tt Can help to ith Cwtumen Seqmendation, tsaud detection
asesment, psonalized tieketng e tcnd analysis. by
bants asn nafe
andlyzing Aage volmes oO) cst customen dala,
intormed decietons, frmpsove Cestumtn enperierces s enchanco
Dvnall ppoutienal 7tcienq
 Seqmentaton ; Stgmant customens bas td
on thein ftrantial
wtm
behauíov prtaencts . thrs helps (n tangtttd manketing Poduet
k7eings &fnpe cwstomen enpeoúences,
t aids (n fndetiying potentral rísks by
Pisk Management i
istonieal data, pa ttena G thnd Tt enables predrct
nalzng
and manage loan povbfulios
yoYe

traud Det eet'en'- Analyze tiansaclion pattens G behauiouns it

identiby Lunuslal activítes that míatt fnditate faud. This
minimi2es doss es alue bo aunatthozed thasatiens.

F) anket Analysís Data Minng amiske in analyzinq masktt ends

predrttng tntenest rate change and undestandrnq the fmpatt
o- econo míe facto xs on the induiby.
5) Pesonalízcd Ceutci 1hough dothre bants Can stho pevsonalizad
tinancial adurce , setivement plannfng bantd on fndr vrdeual
Ctonne prokles.
6) Cedit Scortng Chreate mone accuwrale e tai cve drt fcoving madele,
Constdeinq olde mange oy faetos Geyond Badttional Gredit
Wstong
) operateral eticny - tHelps bane beamlire opnatiens b

fnestíciencíes.
8) Psedictíve
roalysis' By ralqzing hctoñe dlata bants can
he matet conditíons lvtnest ratos predtct
enaldlna betes decíelen motfng.