Saniivani College of Engineering, Koparga on 4216O3

Dep artment of Computer Enginee ring

Practice Questions
Subject: Data Mining and W arehousing (CO 4O3), Class: B.Tech,
SEM: VII A.y=2OZZ-23
Unit I: Introduction to Data Mini ng
Fractice s
Sr.
Question CO BTL
No.
1 What is data mining? I n your answer, address the following:
1 3
[a) Is it another hype?
qb) I: it a simple transformation or application of technology
developed from databases, statistics, machine learning, and
pattern recognition?
Ic] We h a VC p resented a VI EW tha t da ta mining is the result of the
evol utio n o f d'ata base tech n
) Do yo u think tha t d ata m ining IS als o the resu It of th e
evo uti on o f L 3
Ma'chi ne lea rn ng research ? Can you pres e n t S u ch views
b ased o n
the h isto rical pr,o gress o f this dis cl p li n e ? Address the same
for
the fi el ds o f sta ti,sfics and n itio n
3. D es cn be th e step S nvolve d in da ta mlnl ng wh
e n viewed AS a 1. 3
cess of know d
4. How is a data warehouse differ ent fro m a database? How
are they 1_ 3
similar?
5 D efi ne each of the fo IIo WI n8 data mt nrng
functio naliti es: 1. 3
characteri zati o n, dis c ri mi na ti o n, asso cta ti on and CO rrela
ti on
an al ysis classi fi CA ti ofl, regr'ess IO I1, clusteri ng, a n d o utli er
a nalys is. Gi VE exam p CS of ea ch d ata m in in o functionali
b ty us ing a
real- ife da ta base th at a re fa mil iar with.
6. Presen t a n exam ple wh ere da ta mini ng is C rucial to th e s CCESS
u of 1. 3
a bus Iness. Wh at da ta m tn rng
functi,ona liti,ES d o es thi S b US INCSS
need i e,'b)'t
(, think
of th e ki n ds of pa ttern S that co ul d b e ml n e dl 2
Can s u ch patte rns b e ge nera ted a Ite rn a ri Vely by da
ta q u e ry
rocess ln o r sim e sta tistical s?
7 Explai n th e di ffer'enc e a n d si mil a rity between d ls crlml n ati o n
and L
classi fi cati o n, b etween characterizati on a nd cl u steri ng,
a nd
betwee n classi ficati on an d on.
B. o utliers a re o fte n di sca rded as n o is e. H owever o ne p e rso n s L 3
gar bage co ul d be anoth er',s tr easure. F or exam ple, exce ptio
ns in
credit card tr'ansa ctio n S can help us detect the fraudu e n t us e o f
credi t CA rds U S ing frau d ul en CE detectio n as a n exam pl e, pro p OS
e
two metho ds that ca n be u S ed to detect outli ers and discus S
whi ch o n e IS m or e re li a bl e,
9. Wh at a re the maj or chall e nges o f mt nl ng a h uge am ou nt of data
L 3
( eo b lli ons of tup es ) in CO paris on wl rh mi ni ng
a S m a ll
amo u n t of d ata da ta set of a few h und re d e ?

Prepared by: 1. Dr. A. B.pawar

Approved by: Kshirsagar
2. Dr,K.Vengtesan
cu.-.0)