JI

MMAUNI
VERSI
TY
AGAROCAMPUS
DEPARTEMENTOFCOMPUTERSI
CENCE
STATI
STI
CSANDPROBABI
LITY
GROUPASSI
GNMENT

NAME I
D

1)BI
SRAT TESHOME RU0504/
15

2)BLEN TSEGAYE RU0513/

3)DAGI
M TAMI
RAT RU0563/
15

4)DANI
EL BI
RHANU RU0585/
15

5)DAWI
T ZEBENE RU0597/
15
 1)Whatkeynoti
onborneinmi
ndwheny
ouhear
edaboutt
het
erm
Stat
ist
icsandpr
obabi
li
ty.

 Whenhear i
ngabouttheter
ms" Statist
ics"and
"Pr
obabil
ity
,"akeynoti
ontobearinmi ndi suncert
aint
y.
Bothst
atisti
csandprobabi
li
tydeal wit
hanal y
zingand
under
standinguncer
tai
ntyi
ndifferentway s:

1.Statist
ics:I
nstati
sti
cs,weusedatat omakei nfer
ences
andconcl usi
onsaboutapopulationbasedonasampl e.
Stati
sti
cal methodshelpussummar i
ze,anal yze,and
i
nterpretdatatomakei nfor
meddecisionsorpr edi
cti
ons.
Thekeynot i
oninstati
sti
csist
hatthereisinher ent
vari
abili
tyoruncert
aintyi
ndata,andstati
sticaltechni
ques
helpusquant if
yandmanaget hisuncert
ainty.

2.Pr obabil
it
y:Probabi l
i
t yisthelikelihoodofanev ent
occur r
ing.I
tprovidesamat hemat icalframewor kfor
quant i
fyi
nguncer t
aintyandr andomness.Pr obabil
it
y
theoryhelpsusunder standt helikeli
hoodofdi ff
erent
outcomesi nv ar
ioussi t
uat ions,suchasgamesofchance,
weat herfor
ecasting,orr iskassessment .Thekeynot i
on
i
npr obabili
tyisthatitallowsust oassi gnnumer i
cal
valuest ouncert
ainev ent sandst udyt hei rpr
operti
es.

Inessence, bothstati
sti
csandpr obabi l
ityrev
olvear ound
theconceptofuncer taint
yandpr ovidet oolsfordealing
withthisuncertai
ntyindiff
erentcont exts.Understanding
andappl yi
ngst at
isti
calandprobabili
st i
cconcept scan
helpi
nmaki nginformeddecisions,dr awingvali
d
conclusions,andpredicti
ngoutcomesi nv ari
ousfields
suchassci ence,business,f
inance,andev ery
daylife.

2.Ev
enqui
tesmal
ldat
aset
sar
edi
ff
icul
ttocompr
ehendwi
thoutsome
summar
izat
ion.St
ateknownst
ati
sti
calquant
it
iest
hatcanbehel
pfuli
n
summar
izi
ngdataalongwi
tht
hescaleofmeasurement
.

 Someknownstat
ist
ical
quant
it
iest
hatcanbehel
pful
insummar
izi
ng
dat
aincl
ude:

1.Mean:Theaver
ageofasetofnumber
s.I
tisappr
opr
iat
efor
i
nter
valandrat
ioscal
edat
a.

2.Median:Themi
ddleval
ueinasetofnumberswhentheyar
e
arr
angedin or
der
.Iti
sappropr
iat
eforor
dinal
,i
nter
val
,andr
ati
o
scal
edata.

3.Mode:Themostfr
equentl
yoccurr
ingval
ueinasetofnumbers.I
t
i
sappropri
atef
ornominal
,or
dinal
,i
nterv
al,
andrat
ioscal
edata.

4.Range:
Thedi
ff
erencebetweenthemaximum andmi
nimum val
ues
i
nasetofnumbers.I
tisappropr
iat
efori
nter
val
andrat
ioscal
edata.

5.StandardDevi
ati
on:Ameasureofthedi
spersi
onorspreadofaset
ofnumber sar
oundthemean.I
tisappropr
iat
eforint
erv
alandrat
io
scal
edat a.

6.I
nterquarti
leRange( I
QR):Therangeofthemiddle50%ofasetof
numbers,calculat
edasthediff
erencebetweent
het hi
rdquart
il
e(Q3)
andthefirstquart
il
e(Q1).I
tisappropr
iat
eforor
dinal,
int
erval
,and
rat
ioscaledata.

Thesestati
sti
calquanti
ti
eshelpt
oprovi
deasummar yoft
hecent
ral
tendency
,vari
abil
it
y,anddist
ri
but
ionoft
hedat
a,dependi
ngonthe
scaleofmeasurementused.

3.Ber
if
lydescr
ibet
heAppl
icat
ion,
Useandl
imi
tat
ionofSt
ati
sti
cs.

Appl
icat
ionofSt
ati
sti
cs:

St
ati
sti
cshasawi
der
angeofappl
i
cat
ionsacr
ossv
ari
ousf
iel
ds,
incl
udi
ng:

 Busi
nessandEconomics:Busi
nessesusest
ati
sti
cst
oanalyze
markett
rends,
for
ecastsal
es,opt
imizepr
oduct
ionpr
ocesses,
and
 makeinfor
meddeci
sions.Economi
cdataisal
soanaly
zedusi
ng
st
ati
sti
calmet
hodstounderstandt
rendsandpat
ter
ns.

 Heal t
hcare:I
nheal t
hcare,
stati
sti
csar eusedt oanal
yzepati
entdat
a,
conductcl i
nical
tri
als,
studydiseasepatterns,andassesst
he
effecti
venessoft r
eatments.Stat
ist
icalmet hodshel
phealt
hcare
professional
smakeev i
dence-baseddecisi
onsandi mprovepati
ent
outcomes.

 SocialSci
ences:
Stati
sticsplayacrucial r
oleinsocialscience
resear
chbyanalyzi
ngsur veydata,conduct i
ngexper i
ment s,and
studyi
nghumanbehav ior.Sociol
ogist
s, psychol
ogists,andpol i
ti
cal
scient
ist
susestati
sti
cal methodstodr awconcl usi
onsf rom dataand
testhypot
heses.

 Qualit
yContr
ol:I
ndust
ri
esusest at
ist
icstomonitorandimprovet
he
quali
tyofpr
oductsandprocesses.St
atist
ical
processcont
rol
techni
queshelpi
denti
fyvari
ati
onsanddef ect
sinmanufacturi
ng
processest
oensureconsist
entquali
ty.

UseofSt
ati
sti
cs:
 Descripti
veStat
ist
ics:Descri
pti
vestat
ist
icsar eusedtosummari
ze
anddescr i
bedatausingmeasur essuchasmean, median,
mode,
standarddevi
ati
on, andhist
ograms.Thesest ati
sti
cshelpi
n
understandi
ngthechar act
eri
sti
csofadat aset.

Infer
ential
Stati
sti
cs:I
nferenti
alstat
ist
icsareusedtomake
predicti
onsorinf
erencesaboutapopul ationbasedonsampledata.
Techniquessuchashy pothesistesti
ng,regressi
onanal
ysi
s,and
confidenceint
erval
sfallunderinf
erenti
alstati
sti
cs.

 DataVi
suali
zat
ion:St
atisti
csisusedtocreatevisualrepresent
ati
ons
ofdat
a,suchascharts,graphs,anddashboar
ds.Dat av i
suali
zat
ion
hel
psincommuni cat
inginsight
sandt r
endseffecti
vely.

 Pr
edi
cti
veModel
i
ng:
Stat
ist
icsi
susedt
obui
l
dpr
edi
cti
vemodel
sthat
 forecastf
utur
eoutcomesbasedonhistor
ical
dat
a.Techni
quesli
ke
regressi
onanaly
sis,
timeseri
esanal
ysi
s,andmachi
nelear
ningare
commonl yusedforpredi
cti
vemodel
ing.

Li
mit
ati
onsofSt
ati
sti
cs:
 Assumptions:Stat
ist
ical
met hodsoftenrelyoncertai
nassumpt i
ons
aboutthedata,suchasnormal di
str
ibuti
onori ndependenceof
observat
ions.Viol
ati
onoftheseassumpt ionscanleadtoinaccurat
e
resul
ts.

 Sampl
i
ngBias:I
fthesampl eusedforanaly
sisisnotrepr
esentat
ive
oft
hepopul
ati
on,st
atist
icalconcl
usionsmaynotbegener al
i
zable.
Sampl
i
ngbiascanaffectthevali
dit
yofstat
ist
icalf
indi
ngs.

 Causati
onvs.Corr
elati
on:Stati
sti
cscanest abli
shrelat
ionships
betweenvari
abl
es(correl
ati
on),butitcannotprovecausation.
Inf
err
ingcausat
ionfrom st
atist
icalassoci
ationsrequir
esaddi t
ional
evi
denceandstudydesign.

Interpret
ation:St
ati
sti
cal
result
scanbemi sinterpr
etedor
mani pulatedifnotcommunicatedaccur
ately.Understandi
ng
statist
icalconceptsandl
imitat
ionsi
scrucialforproperint
erpr
etat
ion
ofr esult
s.

4.I
dent
if
ythescal
eofmeasur
ementf
ort
hef
oll
owi
ngv
ari
abl
es.
A) Numberofhear
tat
tacks,

B)
.Et
hni
cgr
oup,

C)Jobsat
isf
act
ioni
ndex(
1-5)
,

D)
.Ser
um ur
icaci
d(mg/
100ml
),

E)Cal
endary
ear

F)
.Thenetwagesofagr
oupofwor
ker
s

Thescal
eofmeasur
ementf
ort
hef
oll
owi
ngv
ari
abl
esi
sasf
oll
ows:

A)
.Numberofhear
tat
tacks-Rat
ioScal
e:Thenumberofhear
tat
tacksi
smeasur
edon
arat
ioscalebecausei
thasatr
uezeropoint(
absenceofhear
tattacks)andal
l
owsf or
meaningfulr
ati
ostobecal
cul
ated(
e.g.
,onepersonhavi
ngtwiceasmanyhear tat
tacks
asanother)
.

B).Ethni
cgroup-Nomi nalScal
e:Ethnicgroupismeasuredonanominal
scale,whi
ch
categori
zesi
ndi
vidual
sintodist
inctgroupswithoutanyi
nher
entor
derorranki
ng.

C).Jobsati
sfacti
onindex(
1-5)-Or di
nalScal
e:Thej
obsat
isf
act
ioni
ndexismeasured
onanordinalscal
e,wheret
hev aluesrepr
esentor
der
edcat
egori
esbutdonothave
equali
nter
valsoratruezer
opoi nt.

D).Ser
um uri
caci
d(mg/100ml)-Rat
ioScal
e:Ser
um ur
icacidismeasur
edonar ati
o
scal
easithasatr
uezer
opointandall
owsformeani
ngfulr
atiost
obecalcul
ated.

E).Cal
endaryear-Int
erv
alScal
e:Calendaryeari
smeasuredonani
nter
valscale,
where
theint
erv
alsbetweenconsecut
ivey
ear sareequal
,butt
herei
snotr
uezeropoint(year
0).

F).Thenetwagesofagr
oupofworker
s-Rat
ioScal
e:Thenetwagesofwor ker
sar
e
measuredonarati
oscal
east
heyhaveatr
uezeropoi
nt(absenceofwages)andal
l
ow
formeaningf
ulr
ati
ost
obecal
cul
ated.

5.Thefol
lowi
ngdatarepr
esent
stheamountoft
ime(i
nhours)for80
col
legest
udentsdevot
edtolei
sur
eact
ivi
ti
esdur
ingatypi
calschoolweek:
a)Const
ructagr
oupedf
requencydi
str
ibut
ionof

b)Fi
ndt
hemean,
medi
an,
Modeandst
andar
ddei
vi
ati
onf
rom t
hegr
oupedf
requencyt
abl
e

2324181420242426232116151920221413201927

292238 28 34 322319213116281918122715212516

3017222929182520161117121524252122171815

2120231817151626232211161820231917152010

SOLUTI
ON:
-(a)

Fi
ndcl
ass(
k)=1+3.
322*
log(
80)
=7.
32=7

St
ep1:
Max=38,
Min=10,
sot
her
ange;
R=38-
10=28.

St
ep2:
Iti
sal
readydet
ermi
nedt
oconst
ructaf
requencydi
str
ibut
ionhav
ing7cl
asses

St
ep3:
Classwi
dth:
W =28/
7=4
St
ep4:
Star
ti
ngpoi
nt=10=l
owerl
i
mitoft
hef
ir
stcl
ass.

Andhencet
hel
owercl
assl
i
mit
sbecome:
10141822 2630 34

St
ep5:
Upperl
i
mitoft
hef
ir
stcl
ass=14-
1=13.

Andhencet
heuppercl
assl
i
mit
sbecome:
13172125293337

Cl
assl
i
mit cl
assboundar
y cl
assmar
k(mi
) f
requency(
fi
) cumulati
ve
f
requency

10-
13 9.
5-13.
5 11.
5 5 5

14-
17 13.
5-17.
5 15.
5 18 23

18-
21 17.
5-21.
5 19.
5 22 45

22-
25 21.
5-25.
5 23.
5 17 62

26-
29 25.
5-29.
5 27.
5 9 71

30-
33 29.
5-33.
5 31.
5 3 74

34-
37 33.
5-37.
5 35.
5 1 75

∑7
i=1xi
fi
(
b)
*mean(
x)= = =
7
∑ f
i
i=1
( 5)+(
5*11. 5)+(
18*15. 5)+(
22*19. 5)+(
17*23. 5)+(
9*27. 5)+(
3*31. 1*35.
5)
5+18+22+17+9+3+1
1542.
5
= =20.
5
75

Wn
*
Medi
an(
x)=Lmed+ (-
cf) wher
e Lmed=17.
5 n/
2=40 cf
=23
f2
f
=22
4
Medi
an(
x)=17.
5+ ( 40-
23)
=17.
5+3.
09=20.
59
22
 ∆1
*
Mode(
x)=Lmo+W( ) wher
eLmo=17.
5fmo=22 w=4
∆1+∆2
f
2=17 f
1=18

∆1=f
mo-
f1=22-
18=4
∆2=f
mo-
f2=22-
17=5
4
Mode(
x)=17.
5+4( )=17.
5+16/
9=19.
2
4+5

on= s
2
*st
andar
ddev
iat
i
7 2 7 2
∑i=1(xi
-x) ∑i=1xi-nx2
s2= =
n-1 n-1
s2 =
2 2 2 2 2 2 2
(
11.
5-20.
5) +(15.
5-20.
5) +(
19.
5-20.
5) +(
23.
5-20.
5) +(
27.
5-20.
5) +(
31.
5-20.
5) +(
35.
5-20.
5)
75-
1

s2=6.
9

st
andar
ddev
iat
ion = 6.
9=2.
63
6.Howmany7-
let
ter
wor
dscanbef
ormedusi
ngt
hel
ett
ersoft
hewor
d
BENZENE?
 Wecanuset
heconceptofper
mut
ati
ons,
iti
sthe7l
ett
ers(
B,E,
N,Z,
E,
N,E)t
akenal
lat
once.

Si
ncet
her
ear
erepeat
edl
ett
ersi
nthewor
d“BENZENE”
,Ithas3”
E”and2“
N”.
1(B,
Z)

n!
P(
n;r
1,r
2, rk)=
…,
r
1!r
2!…r
k!
 7! 7! 5040
P(
7;1,
3,
2,
1,
)= = = =420letter
1!
3!2!
1! 3!2! 12
ar
rangement

7.Whati
sthepr
obabi
li
tyt
hat3menand4womenbesel
ect
ed
fr
om agr
oupof7menand10women?
17!
Sol
uti
on:
-Tot
alpossi
blecommi
tt
ee:
n( =(1
s) 7)
= =19448
7 7!

7)10)
7! 10! 7! 10!
LetA=3menand4women,
n( =(
A) ( = * = *
3 4 3!
( 3)
7- ! 4!
( 4)
10- ! 3!4! 4!
6!
=35*
210=7350
n(
A) 7350
P(
A)= = =0.
377
S) 19448
n(
8.Thepr
obabi
li
tyt
hatamar
ri
edmanwat
chesacer
tai
nTVshowi
s0.
4andt
he
probabi
li
tythatamar ri
edwomanwat chestheshowis0.5.Thepr
obabili
tythat
amanwat chestheshowgi venthathi
swifedoesis0.7.Fi
ndtheprobabi
li
tythat
A.Amar ri
edcouplebot hwat cht
heshow
B.Awi f
ewat chesifherhusbanddoes
C.Atleastonepersonofamar ri
edcouplewatchest
heshow

Sol
uti
on:
-gi
ven

P(
M)=O.
4

P(
W)=0.
5

P(
M/W)
=0.
7

A) P(
MnW)
=P(
M)*
p(W/
M)but P(
W/M)
=P(
M/W)

P(
MnW)
=P(
M)*
P(M/
W)

P(
MnW)
=0.
4*0.
7=0.
28
B) P(
W/M)
=P(
MnW)
/P(
M)

P(
W/M)
=0.
28*
0.4=0.
7

C) P(
atl
east
)=P(
M)+P(
W)-
P(MnW)

=0.
4+0.
5-0.
28

P(
atl
east
)=0.
62

9.Inacertai
nindust
rialfaci
li
ty,
acci
dent
soccuri
nfr
equent
ly.Iti
sknownt
hat
the
probabi
li
tyofanaccidentonanygi v
endayi
s0.005andaccidentsar
e
i
ndependentofeachot her
.
A.Whatist
heprobabi
li
tythati
nanygiv
enper
iodof400daysther
ewi
ll
be
anacci
dentononeday?
B.Whatist
heprobabi
l
itythatt
her
eareatmostthr
eedayswit
hanacci
dent
?
C.Fi
ndtheexpect
edvalueandvar
iance

Sol
uti
on:
-letacci
dentongi
venday
=A

P(
A)=0.
005

P(
noA)
=1-
0.005=0.
995

NOofday
sn=400
(
n-1)
A)P(
X=1)
=nC1*
P(A)
*P(
noA)
(
400-
1)
P(
X=1)
=400C1*
0.005*
(0.
995)
(
399)
P(
X=1)
=400*
0.005*
(0.
995)

P(
X=1)
=0.
181

B) weneedt
ocal
cul
atecumul
ati
vepr
obabi
l
ityofhav
ing0,
1,
2,
3acci
dent
sin400day
s.

P( =P(
X≤3) X=0)+P(
X=1)+P(
X=2)+P(
X=3)
(
400) (
399) 2 (
398)
P(
X≤3)
=(0.995) +400* 0.005*(
0.995) +400C2*
(0.
005)*(
0.995) +400C3*
3 (397)
(
0.005)* (
0.995)

P(
X≤3)=0.
999

C) E(
X)=n*
P(A)=400*
0.005
 E(
X)=2

Var
(X)=n*
P(A)
*P(
noA)=400*
0.005*
0.995

Var
(X)
=1.
99

10.Ashi
pmentof20si mil
arlaptopcomputer
stoaretai
loutletcont
ains3t hat
aredef
ecti
ve.Ifaschoolmakesar andom pur
chaseof2ofthesecomput ers,
fi
ndtheprobabil
it
ydist
ri
buti
onf ort
henumberofdefecti
ves.Checkthatf(
x)
defi
nesapdf;
Sol
uti
on:
-.

Gi
ven:n=2 x=0,
1,
2(Noofdef
ect
ive)

3
P= (pr
obabi
l
ityofsel
ect
ingdef
ect
ive)
20
n
P( ()
X=x)= *px*(
x
1-
n-
p)
x

Forx=0;

2 30 3 2-0
()
X=0)= *( )*(
P(
0 20
1- )
20

172 289
P(
X=0)
=1*
1*( )=
20 400

Forx=1;

2 31 3 2-1
P( ()
X=1)= *( )*(
1 20
1- )
20

3 17 51
P(
X=1)
=2* * =
20 20 200

Forx=2;

2 32 3 2-2
P( ()
X=2)= *( )*(
2 20
1- )
20

32 9
P(
X=2)
=1*
( )*1=
20 400

Thus,
thepr
obabi
l
itydi
str
ibut
ionofXi
s
 x 0 1 2

f
(x) 289 51 9
400 200 400

11)LetXbeacont
inuousr
andom v
ari
abl
ewi
thpr
obabi
li
tydi
str
ibut
ion

{
2(
x+1)
,-1<x<2,
F(
x)= 9
0,el
sewher
e

Fi
ndt
hepr
obabi
li
tydi
str
ibut
ionoft
her
andom v
ari eY=x
abl 2

Sol
uti
on:
- F(
y) Y≤y
=p( )=p(x2≤y -y≤x≤ y)
)=p(

y
F(
y)=∫
-yf
(x)
dx

y 2(
x+1)
F(
y)=∫
-y dx
9
1y
F(
y)= ∫
-y2x+2dx
9
1 2 y
F(
y)= [(
x +2x)
]
-y
9
4
F(
y)=
9y

Fi
ndt
hePDFofY,
wedi
ff
erent
iat
etheCDF

d d 4 2
f
(y)
= F(y
)= ( ) =
dy dy9 y 9 y
Ther
efor
e,t
hepr
obabi
l
itydi
str
ibut
ionoft
her
andom v
ari eY=x2i
abl s:
 {
2
,0<y<4
F(
y)=9 y
0,
el
sewher
e

12.Acer t
aintypeofstor
agebatt
erylasts,onav er
age,3.
0y earswithastandar
d
dev i
ati
onof0.5y ear
.Assumingthatbatteryli
feisnor
mallydistr
ibuted,
findt
he
probabi
li
tythatagivenbatt
erywi
lllastl
esst han2.3year
s.

Gi
ven:
-av
eragebat
ter
yli
fe,
µ=3.
0year

St
andar
ddev
iat
ion δ=0.
5year
s

Val
ueofi
nter
est
,x=2.
3year
s

St
ep1:
cal
cul
atet
hez-
scor
e.

x-
µ
Z=
δ
2.
3-3.
0
Z=
0.5
-0.
7
Z= =-
1.4
0.5
St
ep2:
lookupt
hez-
scor
einast
andar
dnor
mal
tabl
eorusecal
cul
ate

Az-scor
eof-
1.4gi
vesusapr
obabi
l
ityof appr
oxi
mat
ely0.
0808or8.
08%

Ther
efor
ethepr
obabi
l
ityt
hatbat
ter
ywi
l
llastl
esst
han2.
3year
sis8.
08%

13.Fi
ndt
heexpect
edv
alueoft
her
andom v
ari
abl
eXwi
tht
heCDFof
F(
X)=x3,
0<X<1

Sol
uti
on:
-

d d 3
F(
x)= F(
X)= (x)=3x2 wher
e0<x<1
dx dx
1
E[
X]=∫
0xf
( dx
x)
1
E[
X]=∫
0x( dx
3x2)
13
X]=3∫
E[ 0x dx
 x41
E[
X]=3[ ]
40

3
E[
X]=
4