Probability and Random Variables 1.

43
SpecialProbability Distributions
Discrete Distributions
Bernoulli trial:

namonlli trial is an experiment that results in one of the two outcomes: success and
failure.

1. Binomial Distribution:
Suppose that we conduct n independent Bernoulli trials. Let p denote the probability of
RvVeORsRin atrial and let X denote the number of successes in such trials. Then we say
that X is a binomial variate and its pmf is given by

P[X = a = nCpgt T = 0, 1, .., n; where q =1-p.

The real numbers n andp are called the parameters ofthe distribution.
hparae
Derivation of the MGF, Mean and Variance of a Binomial Distribution:
Solution:

Mgf = Mx (t) = E<e]

=)ePX =al
C=0

T=0

=n'zpe')qi
T=0 neitaiaiG leimorie no moldor
=(g+pe')"olneod
(Since: When n is a positive integer,
(a+ b)" = a + ncab+....+6
= ncoa-'B + nca-b+.... + ncha"-"b"

-nc,ba"-)
T=0
 andNumberTheory
1.44
Pohability odiieldyldodoe
Mean -ELX
anoliuditela

p)"-p}=n.lp=np
- (n(g +
noidil

t=0

peyn-l +e'(n - 1)(+pe')pe|)},

-fnple'(a +
+p)"pl}
={nplg + p)n- + (n- 1)(q
=npl1 + (n - 1)pl
= np[l + np - p
(1-p)]onoleVbnen8oll36M
=nplnp +
= np[np + q)
=n'p +npq

Therefore,

Var(X) = E(X²) - (E(x))?

=pp + npq - (np
= npq

Problems on Binomial Distribution:

1. Two dice are thrown 120 times. Find the average number of times in which the number
on the first die exceeds number on the second die.
Solution:
n(S) = 36
A=event that the number on the first die exceeds the number on the second die nu
throw of two dice
Then,
 Probability and Random Variables 1.45

A = {(2,1), (3,1), (3, 2),(4, 1), (4,2), (4, 3), (5, 1),(5,22), (5, 3),
(6.2), (6, 3), (6, 4), (6, 5)}
5,4), (6,1),
Therefore, 15 5
n(A) = 12P
36
Hence, the average number of times A occurs when n = 120 is

= Np, the mean of a binomial variate

5

-120 = 50

)A machine manufacturing SCrews is known to produce 5% defective. In a random

sample of 15screws, what is the probability that there are

(a) exactly 3 defectives (c) at most 3 defectives?

(b) at least 3 defectives

Solution:
Let X = number of defective screws
Then X is abinomial variate with parameters
5 1
n= 15,p=
100 20
Therefore,
1 19
q=1-p= 1
20 20
(a) Probability that exactly 3 defectives is
= P[X = 3
12

= 15Cap°g?=455 =0.0307

(b) Probability that there are at least 3 difectives is

= P(X >3)
=1-{P[X = 0] + P[X = 1]+ P[X = 2]}
=1-{15c0p'ql5 +15c1pq +15cgpg}
=1-q {a + 15pq +105p}
=1
G)-(G)G)
=1-0.5133{0.9025 +0.7125 +0.2625}
106

=1-0.5133(1.8775) =0.0363
 Theory
1.46 Probability and Number
i s L
at most 3 defectives
(c) Probability that there are
-PX< 3)
P[X = 3
- (P[X =0] + P[X =1] + P[X=2)) +
-0.5133(1.8775)+0.0307
=0.9637 + 0 . 0 3 0 7 h iio

=0.9944

1
How many times he must
3. Theprobability that a man shooting a target is fire so tha,
wST2lussrn
than
probability of hitting the target at least once is more
Solution:
Let ndenote the required number of times.
Let X = number of times hitting the target
Then
1
P=,and q =1-p=l 3
4 4
To find n so that
2
PX >1]> 3
’ 1- P[X = 0] > /3
1
’ PX =0 <
1
’ ncopq<
3
1

’ n ln

’ n(-0.2877) < -1.0986

1.0986
=3.819
0.2877
’ n = 4
Out of 800 families with 4children each, how many families would be expectedtoha
(a) 2 boys and 2 girls (b) at least one
boy (c) children of both sex?
Solution:
Let
 Probability and Random Variables 147

childrens
I=number of boys in afamily with 4
Then X is a binomial variate with parameters
n=44and p -
;
Therefore, 1
1_11
q=l-p=l-;-5
(a) Probability that in afamily there are 2boys and 2girls ishen
- P|X = 2)
=4c2

8E
Therefore, the expected number of families, out of 800 families, with 2 boys and
2girlsishte
=800 x PX =2) oolo
3
= 800 x = 300
8
(b) Probability that in a family there is at least one boy is
=P[X > 1]
=1- P[X =0|
=1- 4c0pg
1 15
=1-(6)=1h 16 16
Therefore, the expected number of families with at least one boy is
=800 xP[X > 1]
15
=800 X
16
=750

(c) Probability that a family has children of both sex is

= P[X = 1] + P[X = 2) + P[X =3]
= Ac1pg° + 4cgp'g + 4csp°g

14 7
16
 Theory
Number children of both sex is
Probabilityand familieswith
1.48 numberof
expected 7
Therefore,the =800X
8 Proble
=700
1. The
Son

fun

2. PoissonDistribution rÀ(A >0), if its pmfiis given by a

parameter,
variate with b
discreterv. Kis a Poisson (c
A e T= 0,1,2, . Sol
P[X =o=
distribution under the Let

isthelimiting case of Binomial assumptions wit

Poisson distribution np=À,finite. (1)
p’0 3.
1. n’0 2. Distribution:
Poisson
Mean and Variance of a
Derivation of the MGE, (2)
Solution: e -,=0, 1, 2, ...
= z=
Let X be a Poisson variate. Then PX

E<e]=)eP[X =]
Mgf = Mxt) = I=0
fe
th
(Ae')e CO

-T=0
! C=0 So
L
=erele =ee'-1)
T
(d

P
=.d-e =
d [d

= A|1 +A| =l+?

’ Var(X) = El(X*)]| - (E(X)) = +À-2 =A
 Probability and Random Variables 1.49

distribution
Problems on Poisson
1. The number of monthly breakdowns of a computer is a random variable having aPois-
con distribution with mean equal to 1.8. Find the probability that this computer will
function for a month
(a) without breakdown
(b) with only one breakdown and
(c) with at least one breakdown.
Solution:
Let X =number of breakdowns of acomputer. ThenX is given as aPoisson variate
with mean = = 1.8

(1) P[no breakdowns ]=PX = 0]

-l-D.(653
e(ao = e el8 =0.1653
0!
(2) P[only one breakdown]=P[X =1]
e-!
1!
=e(1.8) =0.2975
2/ is known that the probability of an item produced by a certain machine will be de
fetive is 0.05. If the produced items are sent to the market in packets of 20, find
the number of packets containing at least exactly and at most 2 defective items in a
consignment of 1000 packets using Poisson distribution.
Solution:
Let X = number of defective items in a packet.
Then, X is a Poisson variate with parameter
A= np =20(0.05) =1
Probability that a packet contains at least 2 defective items is
= P[X >2]
E
=1-{PX =0] + P[X = 1)°iudneilohiermoaD

0!
t1!
+ ) isoi adsk
bas slih iues=1-e7^(1
=1-e(1+1) =1-2e-l
ldeal-2(0.3680)
=1-07360 0.264
 anNsnberTheory containing at least 2
packets,
7.80
Psbiliy

Therefore,the
numer of
otof
pockets,
1000

-1000x 0.264
aef ctive
264
2 defectiveitemsis
exactly
packetcontains - P =2)
Prohabiliythtà

22
0.3680 =0.189
2
defective items is
numberof packets with exactly 2
Therefore, the 0.189
= 1000 >x
= 189
Probability that apacket contains at most 2 defective items is olstriouon:
= P[X < 2)
= P[X = 0) + P[X = 1] + P[X = 2]
e-0 e-! e)2
t1
=e1+t,gnintsiuc og o od
1)
-el(1+1+)
5 5
=e=0.3630) =0.920
Therefore, the number of packets containing at most 2 defective items is
= 1000 x 0.920

3. Geometric Distribution

The geometric r.y. X is used to

If the first success describe the number of trails required to get the irst su
occurs on the g'n trial, then
P|X== Probability thatthe first (x - 1) trials Sresult in failures and
the x" trial
results in success
PX=z=P,geometric
t= 1,2,3, .. .;q
This is the pmf of a
r.v. =1-P,p= Probability P, p= of success
lot (1-93
2pq +pqP- 2pq p(1-)+
ge')43 -(1
sev +2pge2*1 --ge]pe'
ge) -(1
qe)(-qe')) -j(1-ge'
pe'.2(1- )(pe')-
ainiarmosD
no Dt=0clajb
pe'
t=0
d(d
=EX
1-9
1
(1-qet)21
pqe+
pge pe'-
t=0 J get)2 (1-
(-ge')) (pe'pe)- ge) -
get 1- getql-
pet get
+.. (ge'j' (ge'?+ Ei+
ge' -
;eeeenlutiopioH
hegorg (1o)
n=1
n=1
E|e] Mgf
=(t) Mx =
Solution:
Derivation
the of
and Mean MGE,
tribution: Geometric Variance
aof
 andNumnberTheory
1.52
Probability Plta)lta
p'

Therefore.
E(X') - (E(X)2
Var(X) =
1+9
p?
1+q 1

I+q-1

Memoryless property (or) Forgetfulness property of

tribution Geometric di
Statement: If X isageometric rv., then X lacks memory in the sense that
[X>s+ t]
X>s
= P[X >t)

Proof:

P[X >m) = P[X =m +1]+ PX =m+2] + ...

=g"p+q+'p+qtp +...
= g"p[l + 1
q+g'+... =9"P
1
=g"p.=g"
P[X>s+tnX > s PX > s+t
PX > s P(X >s]
qß =' = P[X > t|, by (1)

Problems on Geometric
1. If the
distribution
probability
Jiven trial that an applicant for a
is 0.8, what is the driver's license will pass the road test on
(a) On the 4th trial
probability
that he will finally pass the test
(b) In fewer than 4 trials?
Solution:
 Probability and Random Variables 1.53

passi
Let X=number of trials required to
with p=0.8
Then Xis ageometric variate
0.2
Therefore q=1-p= 1-0.8 =
ise e
Probability that he willpass the road test in 4th trial
= P[X = 4 = p= (0.2)°(0.8) = 0.0064
Probabilitythat he willpass the road test in less than 4 trial is

= P[X < 4
= P[X = 1] + P[X = 2] + P[X =3]
--'p+p+gp
FP+ qp +g»p
- P(1+g+q')
=0.81 +0.2 +(0.2)]
=0.8|1 +0.2 +0.04)pl
=0.8(1.24) =0.992

2. Aand B shoot independently until each hits his own target. The probabilities of their
3 5
hitting at each shot are and 7 respectively. Find the probability that B will require
5
more shots than A.

Solution:
bouol
Let pË = probability that A hitting his targetobesd el ori lnnu
osu
Then p1 = 3/5and g1 = 1-3/5 = 2/5
Let pz = probability that Bhiting his target bao ns o laolheg
5 5 2
Then p2= 7 and q2 = 1-=
Let X = number shots required for A to hit the target for the first time
Let Y = number of shots required for B to hit the target for the first time.
Then X and Y follow geometric distribution.
The probability that B willrequire more shots than Ais
= P[X =1, Y= 2 + P[X =1, Y= 3 +...+
P[X = 2, Y=3] + PX = 2,Y =4+...+ ...+
P(X =r, Y=r+1]+ P[X = r, Y=rt1] +...+...
 NimberTheo
Probabilivand PX=|P[Y=8+ +
1.54
-PX=1PY=2) + 2)P[Y =4)+.t t
= -rlP[Y =r +2)+..t*
3 + PXPX =
2P)=
PIX=
PN=nPy=r+l] +
=redat n
PX
PX=rlPy =r+l) +
...
P[Y =rt2)+

r=l

-'Peap[l +q2 +gåt.

rl

-PlSo)=Plag2 +(q1g2)+(a19) +..]

q1 r=1

- g21 +192 +(g192) +..

1 3 2

FP21-q1e57(1-)
6
31
3. Acoin is tossed until the first head occurs. Assuming that the
tosses are independe:
and the probability of a head occurring is p, find the value of p so that the
that an odd number of tosses required is equal to 0.6. Can you probabilty
theprobability is 0.5 that an odd number of tosses is find a value of pso th
Solution: required?dsiong
Let X = number of toSses
required to get first time head
Then X is a geometricvariate
with
P[X =1]+ P|X =3 + 0.6
’ 'p+p+g'p+..,= 0.6 s ole
1
+..) =0.6
=0.6

(-q)(14to0.6 ’ P(1+o)0.6
 Probability and RandomVariables 1.55

0.6(1+o) =1
0.6 +0.6q= 1
’ 0.6q =0.4
0.4 2
0.6 3
2 1
’p=1-=;
It we take 0.5 instead of 0.6 in (1), we get
1 1
=0.5 =
1+9
’1+q =2
’ q=1’ p=0, an impossibility

Continuous Distributions
1. Uniform or Rectangular distribution:
A
continuous r.v. X is said to have uniform distribution on a, b| if its pdf is given by
1
if a<r <b
f(a) = b-a
0 otherwise
It is used to model events that are equally likely to
occur at any time within a given time
interval.

Derivation of the MGF, Mean and Variance of a Uniform or

bution: Rectangular Distri
Solution:
1
a<a<b
fe) =
0, otherwise toitolet

1 1 [etzb
Mgf =Mx() =E(e*]= |e b-a
-da =
ba t
1 ebt eat
(6- a) t

Mean = E[X =coefficient of tin 1

(b- a)