i ge Re
106 | ENGINEERING MaTHematics-ill
THEORETICAL PROBABILITY DISTRIBUTION
A theoretical distribution is the frequency distribution of certain events in which frequencies are
obtained by mathematical computation.
There are two types of theoretical distributions :
(i) Discrete theoretical distributions.
Gi) Continuous theoretical distributions.
some of the discrete and continuous theoretical distribution are :
Discrete theoretical distributions :
. Binomial distribution.
» Poisson distribution.
. Power series distribution.
. Multinational distribution.
. Negative Binomial distribution.
. Geometric distribution.
. Hyper geometric distribution.
Continuous theoretical distribution :
1. Normal or Gaussian distribution.
2. Rectangular or Uniform distribution.
3. Exponential distribution.
4. Gama distribution.
5. Beta distribution.
6. Weibul distribution.
3.6 BINOMIAL DISTRIBUTION
Binomial distribution is a discrete probability distribution, Let an experiment consisting of 7-trial
be performed and let the occurrence of an event in any trial be called a success and its nor
occurrence a failure. Let p be the probability of success and q be the probability of the failure i
a single trial, when g=1- p, so that p+q=l.
<. Probability of r successes in n-trials
P(r)="C,q"" p', r=0,1,2,.
BAUR
n
or PLX=r]="C.q"" p’,r=0,1,2,..
where P(X=r)or P(r) is called binomial probability distribution of a random variable X.
If an experiment consists of m-trials and this experiment is repeated N times, then _
Binomial frequency distribution or Expected frequency of r successes is Y
* fFO=NPC)
or f(r)=N."C,q""p’, 7 =0,1,2.....
HYPOTHESIS FOR BINOMIAL DISTRIBUTION
1. The variable should be diserete i.e. r=0,1,2,3.4,
2. The happening of events must be either a success
3. The number of trials n should be finite and small.
4, The trials or event must be independent.
5, The trials or events must be repeated under identical conditions.
A
n
etc, and never 1.3, 2:1,3-41ete,
failure. Sranistics | 4
5.8. CHARACTERISTIC OF BINOMIAL DISTRIBUTION
1. It is a discrete which gives the theoretical probabilities.
2. It depends on the parameters p or q and n.
3. The distribution will be symmetric if p = q. It is skew symmetric or a symmetric if p # q-
4. The statistics of the Binomial distribution are mean = np, variance = npq and standard
S deviation = npq. :
5. The mode of the Binomial distribution is equal to the value of random variable X which has
the largest (maximum) frequency.
Example 5.14 : For the Binomial distribution, prove that :
Pr+D= oS ie P(r)
which is known as recurrence relation for the Binomial distribution.
: Solution. By the Binomial distribution, probability of r-successes
PU)= "Cg" pj 1 = Oud Qyvnet. oe , a (1)
= PCH =C,.9"9". Crrorsll .@
From (1) and (2), we have
Per+l) _ nt ght pt rin=n)!
sae ae), (r+D)Mn—r-D!t nig’ p
_rin—W(n-r-D p _ =r) p ‘ 7
“GeDNG=r-Dig CHD :
'
Hence Pen =F PP) pire
..MEAN, VARIANCE AND STANDARD DEVIATION OF BINOMIAL
DISTRIBUTION
We know that
A (1)
(2)
mean. =p,,
__ and Variance =n =<):
- where iy! and pz! are first and second moment about origin, respectively.
We have first moment about origin,
: 1H = PO)
7 08 | Encineerinc Matemarics-III
where P(r)="C,q"" p’ is Binomial probability distribution.
ss it aS r'Cg"y’.
&
meee ++ p p(r)=0 for r =0]
Fant P as
z nl wnt or
=2 Go-pla-ni? ee
ae n(n-I)! ee
“PL Gopnai-r-
C= aoe DS
¢-D'@-1-r-D
amp, "C..g" 7p anpdgt py
oe
a wy =p te q+p=l
Thus, mean =p,’ =np, ie, m= np. 3
New, second moment about origin :
we Se rened eer
2 0 70.
“3 fre rVC ge [oP =r(r-1) +7)
rr-I"C, gpl +35 "Cg"
a
Fo
_ i ile ere
fede Dagopit 2 tP
. [By G
US rr=)n(n-l)(n-2)! 44
St DOD? PP
Oe De
"DPD Gass te Sranistics | 409
Se oes "AC gh FF trip
=n(n—1)p? (q+ py? +np-
= nt p?— np? + np be q+ p=
=n'p +np(l-p)
=> pl entpttmpg ¥ tege-p] (4)
Since, variance = My = (u, yi
=n’ p? +ripq—(npy° ‘ [By (3) & (4)}
= mpg.
~ -Hence, Variance (Hs) ~ npg. (8)
Furhter, standard deviation (S.D.) o =/Variance
or Inpq. : By ©]
Example 8.15 : Ina Binomial distribution the mean and standard deviations are 12 and 2 respectively.
2. Find n and p. ;
Solution. Given: mean =np=12, ~(W
S.D.=npq =2 => npq=4. ?
pq. 4
Wehave, yy “72 ~ . [by (1) and (2)]
aes
> : x
Ly 2
Since qtp=l.=> p=l-q> es = Pz Ans.
Aas.
5 .
Now, form (1) np = 12 => %3=12 => n= 18.
Examplé'5.16 : Using the Binomial distribution, find the probability of roliing at snost 2 sides in §
rolls of adice. »
‘ 7
Sichen. “Glen: 2S. pt : [Probab of raling xis
5 410 | Encmcerinc Matuemanics-Ill
By Binomial distribution: P(r) ="C, q"” p" e
P (at most 2 sixes) = P(r <$2)= P(r=0) + P(r=1) + P= 2)
AT me
JO Oe)
Sh 625
= BGr542)= 7. Ans
Example 5.17 : A coin is tossed 4 times, [Link] the probability to getting:
Two heads,
(ii) at least two heads.
Solution. Given n = 4, let p = prob. of getting a head in a single throw of one coin
tes, p= and q=%4
By Binomial distribution: P(r)"C, q" p’
Wry
(i) P (two heads) = Po=2-'c(2) ()
6-3-0375
16 8
(ii) P (at least two heads) = P(r=2 or3 or 4)
P(r =2)+ P(r =3)+ P(r =4)
-e}-=G))-
] (S}atc(4)} (4) ofl yt i
G3) (3) (3) (3) ‘ 6(3)) +a)
$800 L [4464441] =750. Ans.
(iii) The number of expected families having no girls (i.e., r = 4 boys)
(ays 1
f(r =4) = 800%" <3) (3) = 800x5= = 50. : Ans.
(iv) The expected number of families having at most two girls, (i.e., at least 2 boys)
= f(r =2 or3or 4)=f(r=2)+f(r=3)+ fr =4)
IN fly: yy LVN
=800) ‘c,(4) (1) +o f2) (1) we (2) (2
(3) (3) (3)() + (3) (3)
=800x11 - 550,
16 . Ans.
Example 5.23 : Out of 800 families with 5 children each, how many would you expect to have (i) 3
boys (ii) 5 girls (iii) either 2 or 3 boys 2
Solution. Given that : N= 800, n= 5, and
Nie
1 1
Probability of having a boy, P=>ythen 9=1-p=1-5=
() PG boys out of 5 children) = P(r =3)
ae sie
= = =} =—=0,3125.
ol 2) 2232,
.. The number of expected families = 800 x 0.3125 = 250. Ans.
(ii) P(5 girls out of 5 children) = P(i., r=0. boys)
neriy 1
=c,(1] [=] =—=0.03125.
<(3] 2) 32
. The number of expected families = 800 x 0.03125 =25. Ans.