Data Aralytics e cycle Ouerview

bihad is Date Anautics? i

Data Arautics is the process o etoming data sets
4o onlde acut the indaszmatiorn the Comteins
19aicus teo ls +echniaues mnethods
invoes
to analy ze 4he Data, ident the patenns.
eatact the insights that Can
business decisicon o sclve pncblems

insiqht is Totthing bo meoningtl d selevan+Conclusio

it Corn be disuiptiie, paedictiúe opnesaiptive

Dote Arnaytics Jyecle Ovenvieu

what i5 Data?

coe coil) d'suss 3tenms

To del ine duta 4 ils inpetance
) Measwneme t
Dada

Nniable : is a chanack istics any emtty cohich that

tekiny on diHeort úalues.
oroe ss is wsed tð
Measuere ot - 4 is ohen a td pce

meanenent:
ane called a
 meausememt

Dataane e coNded
Data -

di erent ays
qenenated
Data Caa be mdhines ¢human
chme onbine

hemans,

pnoblemsonng
that
qenenaing Data is not a impont nt
Aata specille onmat is
rom anias saunce J
Soolat
aen geting the
that i4 is stone

Now cohy pata is imortart?

pat helps in making hetter decisiD
The
helps in sove pnoblem by indingAeasan
wnden pendehmance
what is Data Analytics 2
Date
IEis the sciertilenoces tansom
beten Decialon

Hhy aralytics is impetan?

The oppotunity abonds ten the se d amautics
big data such as i
fordeermning
neo
Cheditsk
medici nesespeclaly in nealth
Developimg
e industay
Find out pessible cybenlmes
Retaiing most halable Ustomens
 Page No.:
YOUVA
Date:

what is Data Anaysis

JAis a pnoce ss exaning tomsjaing aanging
Sac data iina speitic coay to qemenate usell
niemation tlomi
It Simpy examimes the pa cohat has happened
im the past.

Data Aralytics NS ata Aralysis.

Data analyics /s sdudy ing akout cohat

happen nttene
USing cata
eNents
aralytics Can pred'ct possible

Fics
Arayss, may be qualitative an quantitetive.
Oualitatve eans decisiom masHy lbased on
intutions.
quntitatve
oth the hela teamtulas
make tae decisioms.
Data NiSualization

+Into is the preess, pnocess o Conventg

date nea undenstood-picnes pictnil
oms
Ceasiy
 ile ycle ovenvjeco
f
Data Analytics

deines the noadmap of

ileeyce
Dacta Aaltics qenatd, used.
calected. Processed
-coay olata is busiess qoals
malyzl to acioroplish
cinculaenesentatisn he
Analyst use the
to analyze data inafaroand backo.
lide cyce

obective

Data Ondensta
Visualization
Dadg Aralyties Dat

amattics Dato
Cleaning ¢
tramsleat"
Dat
enkancemeat

TA frovdes sysknatic method

to cmuert t tpn elat
into ine mationo that Canb
to achieve
Ogpnizoional pnaject goals.
 YOUVA
Date:

key Role do
a sucess7al Araltics t
[hene ase centain key oles that ane nequined to
Completed Aralyticat team to enete projet s on
amalities Success uly.
e key oles e

Business Usen -

he usiness. Useis ahthene coho undesstenmd

he mun seo
This usen qives the aduice
Onthe poiect d eplain how the operations om
he olp ane done
Business Mamagen. One'manáqendeep subject
maten expent tuill his ole

It is the One coho is sesponsible te initiccde the

froject
He prouides actual a quine ment anhe pojec
psesent asic business issues
He urds f
qereaiy provides coñat paojeti meecds to
pnovides the
Tois pens of.
ochieve
his penson imntnedte çlaniy cohat the inal
nesut shoudd be.
 M T

Page No.:

Date:

3)

Nesponaible onn ersuning

This penson isi5 nesponsible
milestones ¢ 4he ouesall puspose the
key net omtíme with he expece
phoect e
quality stomdands
4) Business Jntelligene Analyst -8
This penson Aocvides business comain exper
d deep undenstandig cdata key entu
iandiatas Ckpis) kes natnices ausiness
ineltignce nom a nepaating en spective
They typially cnease s nepasts¢ dashboard
Ahane knousledge doda souxes.
S) Doabase Adonilstoato CO BA)

DBA lacilitates4 ananes the dadabase envlacm

to suppont the annalytics meed ss qthe teanm
pensenprovides alcesS to keu databases ¢
tables
ensunes that the appnepniat
Seinity is provided an
Data Ena'neenst
person coho grasp+the deep technial Skills
assist partiulanly im oenking orth datatases
ane skilled in coniting'¢ optimizlg
TheySaLquenies to esicierty manage
extat daa Lomdadabases,
 M WT F S S

Page No.:
YOUVA
Date:

penson sppart on doct

provides doc a mtak e into
the analytic Sandb Cie oonk spoce )
Dat Scient isA

This pensbn acilitaBe onth the subiject matten

expentiese jan 'aalyical teckmiques, data modeling
Aapplyig caecanalytical techrique s tna
qiv en issues.
He ensueS Oyerall amalytical obietives
met

+2 PHASE -Ii Disoves

This phase is all aboud definingthedata's pupase.
i tooto ochieve Aby tthe endoi yck
This, phae identi cal chjective.
uning this process tean leanns abact toe
bustss
businSS domnains e check ohetthen agan
-7atien preuiously coank on Similas yaoyect
to neen to any aaaning
The eorn also enalua the: Aeceonolog y people
data ime inthis phase
eanning the usiness Domaun &

Undenctandirg demain nea of the prcblen s

essemtial
mary ases data scientist oill havs deep
Onptatoral d quantHatIve Knocolede 4had
ehsay applied om tuciness
Date Scienst have deep Krooledqe a methads,
hembuestoq
busin
 M T W T

| Page No.

Date.

At Ahis canly pnoCess the te

Shgem the
need s to deen mine how nuch
domain kncoledqe 4he data Tieotst Meeds
to dexelop modes in phase
2)
Resoues 4naming te probem
pat o iscoveny phase 4he team needs
to assess the avalable to suppa
neSocnces
the pnaject
Desouaces lke lecholoqy tenls systenms
data dpeople
FIndot cuhat ty pe skills danale s eulllbe
nequined that may mot exSist taday
tonong tegoSucess ewhat types skis
daoles e negkiaed to deAelap the aojec
Eindoet equisik devel o eepent)es
anagnizad
need to CLtivaed
teke imenttoAy types date availabe
to the tea
Cons'iden.
tthepjec
dota is Qvulable
sutticitnt Go Suppatspaoject's qaals
teao must heed to Colect additionaldata
eithen Arthase H nom Out Side Souae
tsa
i4 cdadta
existing
is less ha
doada.
sequinememt
sine4 sLope oy the poaject is Me duced
 Page No.:
YOUVA
Date:

3 1dentiyg KeyStack holdens. 2

Stecknoldens
A+ this pomt it is best poactice to cenite docwn
proble Staemeant share it coith Stakeho lders.
sake holden is a pes idiiduaqroupo organlzat
that has a nsed intrest in the successlailune an
outone
a project , busine ss on oncanizatio )
(eg custome, Empbyees Isestans,suppliess etc
tea memben as
ras may hean slightly diysn
AhimgsdSo have Some isha d'ieert ideas a
possible soluions
tan these neason t+ leads to onalytical pdon
Nence t is essemtial Aha dea
cleanly antiuate the Cunend siadicn
mon chaemges
As apat this oiity Ais impasta to
ideni he mam cojectives the pnaject
Additioraly Conaiden dojective 4d the sucess
Cnitenia pn te pnojec
J4 is begt pracice to shane the stadoment
apals Success oiteria coth teom-membens
Aqongoith shane ht eulane caltenia teo
Sucressitenia mot'aes theteam to planth
phoject poec 4eom will Reach its goals.

ailune ctenia coll quide the te

co hen tis best do steap tening to SetR
 Page No.
Date.

Irdenvletoing the inatics Spnsan

The dean Shuld plonto calab¡nate co
stakeh kdens o daty &toone the
Anobles. annaluti
4 poel sponans hayhave
does no meet the desined pedeined
otccme thenir
his Cose fro tean
Cots ust uSe
knowleqed expenties to the
dateSiemce tecon
kngcaledqe to typicaley
yndenstand the have mane
the stke roldeS aob lem har
Cantake hemce slate
cbiective sdenceea
apprsch based
theln eenlene
ne possible ay to
the proiect chncuemt
Sponson to hS 1ssus
ocis
while nCleaaly de
date sciemce
fecus on de
<oals. onethods needed team
achive
hhen
teai ineniewmg
meec s te the main
eke time stake oldens k
froect ponsans bcoz to intéuie the
oho tends to
4umdingpajetspönsan
tne to be
pojec sponan ips tpnindenietsir the
pepaae tate
quemsions
nevle
 Hypotteses- isthe asdanption rade obeud the, erohem
TF S s
Page No.:
YOUVA
Date:

s) Developing imitial hypotheses identibying

Sounces 8-
polentla Data

discovesy
Deseloping IHs toes |sa key fact oy the deas
Th'is step iolve S anming thod
phase
he team Cantest oth dada

These IHs am tthe basis o4he arytical

4est team oi "ak
5e t in iatee phases

thothe pand d thls piocess s to involve

qathenirg 4 assessin hypothesis7ram
who may
`Bakeholdens 4 domain expets
have te Solutions d Cangiesuggestioms on
ideas to test

the
discovey phase identt
need o Solve
dote 4he 4e0mm Cojll
pnobemn Considen volume Aype s & tinme
span o the date needec to tes t hypothe ses.

The eOn showd pentomS

doning the lscovey phase

Hdentity Data Sonce

ata SOnceS
Reiew th mate a list o
nake invemtony b datese cunemty
ayailable
Makea da jmvemtoay dloda set taat
bmbe punchased
 M
Page No.:
Date:

opte aggoegaJe data Sou

Revieo the mo data
4) Evaluate the data Stuttnes 4 teols

PHASE- M - DA TA PREPARATToN.

The Second step og Data aralytics is the date

ipeaation ohich incudes ethe steps to explo
pnepcess 4 onditien doB prioIto mo odellir
aalysis
hs pbase the tean needs toOneate
eminern ment tually this is done
aepaning Onalics Sanbox
e get th Analtis
Sandbe is othing but
analys date siendls olate
desh 4 yaidadelgo dengreens Con [Link]
pithouct nithsmodel d hypthe ses
ectingpooduction syslean an date
pente ETLT pnoce ss, trthedt
into sqndho ne he clata toanslam
\s
Loaddata
imto Somolbex he

he 4eam qmay penanm

hel team mem bens to data wiSualizatin to
um dénstad date
includ'ing tstnerds.icutlens4
date aable nelatioai
 M T W T F SS

Page No.:
YOUVA
Date:

The st Subphase o data pnepoatien nequines

cbtain an graly tics Sdrolbax
ncobich teammcan explone the dota thot
coth l'Ne pon cduction databases
ineneni
dereloping the
He Cnaly ftics sandboz. 4i a
itis
Ahemdereloping
tocalect allkinds dota stone
best pnactice
So that teom Can hamdle hgh volume a
data fan big aralytics project.
Hhis n nclude evezything pncm Sunany lenel
Qagneaated data, stnuckened dateSacw d a a
aeinstnucteed data edc

isto
oten the mimdset e 1 paups preu'de accesS

pnovide the mimimm amnt date nequined to

allo the tegm to acheiv e the ojecties.
anvensey the date Science t e a c o c S to

ocess CNEnythig
bete to handle
mNe data iS
the pnpose niven
oppncaches to test a aiety o ideas
Vie co ace ssg date
Be cau Se tris dibening
date
Science teamto Callaboate
ial
to atonplish sdls
 ETLI 8
usenspeteneiaact ta
mfa

ans
ate
the
In ETL extnact
prcesS fs to
Load doto stof apaech
dLoad
Hinto
Qnalutic
Sand bosk
Heoeven the
d
in ts
duens slightly dos is
extoacted

Sathis Case the dota sto-o

aralyst Cor choo se leave

whene Gtate

iHeoed
ETL, Outliens Tnay be
trarsered ddeancd beteone being
Joaded into Jda stone

the EtL oppnoch qves he teary

o clearr data o anayse the te to
en tte Jota has r0s beerr, Joad i t o tthe
La'ises oess te the data in iHs otqiral
dirg hidders
t
Ahe rolsqit srrdaot (on quckiy

weainingt
 M

Page No.:
YOUVA
Date:

Step w)thin +he Data Aralutics

Date Condikioning nuolve

Cam nvel. many Conple Steps
mengi datasels 0n othenwise qet cadasels
nto a state tha enables aralsls inunthenplases.

Data Gnditining set is penlomed by amly 11,

fants o ths phase inuolues dectdng chch aspecds

t paiaan chtasets
dotasets col be Lad to aralyze
Bolater steps

the team has Collect ed d ootained. same

B datoset that needed pos -aamalysis
We dak. 'sualization
-asetl Sep /s o

dte uisualieat^ teol

4eol eaMnes the date aualuty
Sucha coheHaen t e date Coscteuns. my umexpec
othen diny data.
examne) he skecnees (e (t is a meoss
the distoibt
nea) valued nandonn ianlables
Review dato onsune tha aleution nemaunec
nsistet coithin columnSoacoss tab l e a n
a gie data, ield.
 OnSste

Stey
d i s t a i b u t i c o n

clata p o p l a t i a n

HDees
the
2 the
nepnesemt
date
a l the
the cada ase the
Dees Vaniabes
iotenest 2 anÇ
~nelated monthy?
timt cweekly ?
ten daly
2 a s u N e m e r t clialy
e d / n a m m a l i s e d

Standandiz
data
1s he Phepatiam

Tools yon he Data

Common
phase g

Hodoap massively paalel

Can peapnm
qenomic
pansingGaps location aralsis
-analysisde.

Raasldes Gur on cAeatin

ralytie oanklous ncuding daae raripulatia
4 senies e analtic evest Suchas
dota oinisg echiques On postgaes SoL4
othen Big data Sounces

8
IÀ
J} iS c nee
opem Sowce pawe
kon coonkg colth messy date
pate aqleA is a imdectve
data cleaming dnmsohmation
 Page NO
YOUVA
Date

PHASE -3.
Model pLanring
science 4ean identiie s
Ln phse 3. the date Clustening classiyng
models to cpply
cpply do
to the
+he date lon
nelalienships inthe on
dte deperaing
on inding
the qo pnajecl:
this phate he teamm netereS hypatheses.
helps he t eam
hchis developed in ph -I
pt4.&
the analytics to exeue vein he
select nht methed to achie qeod.
Explanatian dd anlab le Selection B
Data
takes place im the dat prepaatr
Data Exploat ion
pha se explanatian
phase 3,the cbËective o dota
toundenstend the nelotionships amona the
is Selectio t e vàniabes
Naiables to
&methods to mdenstamd cessai pae blean domo

As anlien phase s et the data analytic

Co ith
time 4 tecs
lecycle. it is imetant to spemdcOank to nake
att ention on this pnepanatony coank to
phases al model selectian ere
cut!
Subsequent
easien

Commmon away to Cncuet this Steyo invoe

A Aote ylsualizaten
tols to

date explenotlon nthisou

Aperectig the
team in paevlelnghedota d accessingnelatlens
 Data zima " duning pilot phae.
W
Page No.:

1odel Stcleclon

Jn this phas subphae, te leam's

to checse an amaltical echniaue
ord qo ad o

mìoves
anea âm ton dealng with Big data
dederminirg il the tean oll be
echigues thot are
Suitedlon sau eue
usin
a hyald appech.
Lostly the deam should
Ond
documertt he nodel ake cane to identle
making
madel.
ossmptnns jt S.
chooses4 Constaucts poelmi
wes SAS
mode Cheation
ALthoughd these dsls ane
rnd mochint designed Jan datemi
toey nau have Jeaming algqanitm d
odes to vaay mtim cohem dotamin
dota. lange dadeses applig the
As Such de ann
these aganthm Soitt my Coosiden nedesi
to nun in
pilot phase
ngpilot the Datebase
he tfam
Con moNe to tthe
andel
Kinaoledae te
 3hS/AcCE S heps Comed dienntsyslerms Jote Sowrey
M WTFS S
Page No.:

SAS - Setisticad Analis Sskm Date:

YOUVA

tools tan themodel plonnimg

phase
R Jt as a om plete se dmodelig Capabltr
oith high-ualithy Code
inentere coth
14 has ablity to imBentoure cuith DBS Nla oDBc

Data la a n opem Sowce Coneon

R Conteins soo0 packages to Data oraysis 4
gnaphiwl eaesentet isn
D SAs/AccESS - paovide s mtegation betcEen
LSAS < amaly ics Sandbo ld nutipie data
Cannectos Such as ODBc,JOßC oLE DB

Cxtaacts
hut coth SAs| ACCESS
Aelatinal dodabases

PHASE 8-4 onEL BUILDING

Iotis phase date sScience team build

datases en thaining, Atng pooduio
punposes
These dotasets enables the clata sclentist
derelaped tratait. the aralytica! madel d t

An amautical 3nodel is cleveloped &tomt

data enalated ayaist he test dhe.
 M
Page No.:
Date:

Asa fart oh this prase. the

th
needs to exe(ute dhe models deined. n
Daing this phue WsersSUnndels

analidical siw [Link] sSuch as RD SE

Cnik eta d Smal cdaesels fen testngn
Tn phase 3, he Aeom
teon nayy have had kno
o-nelated yable). cohich coil) bf
an denied once the models oe
atUal eeut
he deteus
model
Gnstauchn madeling 4 oqic
is Coonplet ed. thene 'sns need
o deteils an
Laily engaton
Creatg a models hát ane Scites le toa
neQwses tomeet he model Speci
denloped utimaey tre cpiective being
autind ai
uesoAS to Gnsden
imcludes these
Does the modo
the tet data ? appeon Vaid acumatt
2.

3 gal? the inds

stciently CeAate A0 meet
maclel cho sen Suppot

Onete data science

teagn
evaluae elthen
paeblen
next phase Lta Aralics ue cycte
 T+ is a Grnprehen'ive dote mimirgt meietive aralysis
MT W T
sw, thdelos ade scientisSe-em Page No:
14 podes wn iniend ly mten<ace toer pla tye e vsuE dote.
* Common tools onthe Models Buding P ha

SAS entenpnise mmen alooS cusen toDyn

poedictne 4 desaptrue moclels bas el onJange
dolumme s data om Athe enienpalse
acsoSS the

SPss modelen - yens mettod s to explane a aralue

dota though G0}.
md lab pdesa high-Jexel Jongiae ten
jendeanmig avaieyoq data analysiaalao dota
explotio

Alpine minen prouides

paovdes. ont erd h0g te

* foee open Souxe tocls

PLIR isa-fascedinal language on

fostgress SOL with R using this tperoch R
Conmands Can ekeCutein DB

ctave - taee slo

emputatio al
data
analys
nodelling. has'some atlab
Hnetiaity.
th an analutis
jaee n amimig Slwlcneoted
Jatek h e package

ConeeutedinJa

ytho4Sq.
 M

Page No.:

Date:

Comnumiak
Resu ds.
PHASE
exeution model tem Ompare the
A;te cotela,
atcome ái4h auCce sS lolwSe

pant o his phase tearn needs to dekn

HS a objeCtives.
i succeéded on foiled in HSseenas
shoctd t be
Here lailwe shatbdn be
neqatiwe out comne Istead iH shoud be ui
data doesn t po
a tngeQnough evldenceto ethen accept n
et hypothesis.
The key nembe hene isthat the teomm mus
detenine thot the
the Rhetgh sata eahethen
ohichis outined 'n PHT hypothe
bme 4ime teapmshave onlý
done supenicia
aralusiS 40hichis not abt to accept
Neiet he ypothesis Cthen 4ime team
pendams
ONRk whem Aobsthnaluysis to showthe neu
pesuts ray no oe thene
iis inn partar to balamce be
toee thesi
eing ovenly cnitical f
Oyenly api rristic whe analazing dat
The anaiis should be
to ensuDe that tte Conducted
nesSuds
nesd si aAe
satistYcaly
nalysis is Conpletete its
indimgsho
 Key indgs srothgot mest imp¢ neleyamt,bub
that emengamalysis.
Page No.:
YOUVA
Date:

Whem CoTmmuni ating tthe nesw4S wth te ans

Ais inpantant o be cleond Jousing
jouing
the N e t _ ase mot valid akes the
Ba'ustrent to the madel< itenatively pentan
analysis to Ipnave ds alidity
eb actea nedut the ideas
fomulat ed <aAeCardled in eanlien phase)
The best practie mthis phase s to cOnd
4then Select tte thee
althe jndigs 4them most
Siqni<iánt cnes that Can be Shaned colth the
stalteholdens.

Noco 4earn has nun the model f Gmpleed

though discoueny phase leanñedogaet deal
abou the daaset Considen cohact obstales
cwene in the poject deohat Con be mpeoved
m the
7utine
Stakeho ldenn Oust ndentand how tthe mode
absets thein pnocesses.
As a esut o this phas e the team oil) have
majan einsigbts
deniv ed nom the oaysis
 onS/w tha st
Page No.
Pocesses data imthe CRAM). Date:

pHASE.- Gi Ogesationalize
Tnthe iral phase the team ComomumLCad.
tte beres o the pioject mane proadly
set up a pilbt poolect toto deplby
deplby ima Contt.
toau

Daing he pilot esoje ct tte deorro mayneed

Comsiden executing the alanttm inthe datah
atten tram coinemmenoAy tool Such as R

TeTAnyespecalluo langen Joddsets.

Conside uming the mocel ina pnocict io
Lasioment fan discnete seto
coichtestthe mocdel ina products e Ne seting
Thisalloos tthe teonto make
odustnet Qmy-needed
the modeli
This phase nninqs
ehpeialuy procickiam enqineen team membe
issues <oncenS sho have
paect tem beycrnd Coe
I4 1p'is
beyomd 4he
model to oeanqe
p the eohich
eohih ay
mocnote

Aeined
neAeimed onneu)
pilot Pno le ct is a Smal scale peimmany roject thatis wsed to
M T T FS S

nat

Thene ane main delivenable Cambe coeated to

meet the meedS most stakeholdens
s)Code
2) Techmiad Specs
3)) loematlen 4pn Arasts.
ehesenteatloo ton pooject Sponsas
These keyowpt he key olp toeach stake
holdenham ralytics paeyec 4uhat they
ubunly expect at the Concuso oha poojet
1) Busmess usen - tnies to dedenmimes the
benelts palicle) indngs ina business
questiomS selate
noject Spn son He'asks

business impact the Pooect the sisk

neteunn On inVestmt CROT)

Dedenminesipaaject Conpated

adands do needed to
Business Aralysts
the epontS f cashlboand he mamaq e
Gol be impacted d need to t choge
Dota shane hein Code
Engnee
Cmalyt ics proeet f cneate q technlcal
doceumet on how to mplemert it
Date scientst - Shane tte ocle f esplain the
Stakeholdes
mode to penS ynanagas 4 athen