CE 207

Applied Mathematics for Engineers

Course Teacher: Dr. Md. Hadiuzzaman
Room No. 544, Civil Building
(Lecture 6)
 Normal Distribution
Discovered in 1733 by de Moivre as an approximation to
The binomial distribution when the number of trials is
large. Later derived in 1809 by Gauss.

Importance lies in the Central Limit Theorem, which states

That the sum of a large number of independent random
Abraham de Moivre
Variables (binomial, Poisson, etc.) will approximate a (1667-1754)
normal distribution

Example: Human height is determined by a large number

of factors, both genetic and environmental, which are
additive in their effects. Thus, it follows a normal
distribution.
Karl F. Gauss
(1777-1855)
 Normal Distribution (contd.)
The normal distribution describes the situation in which very
large values are rather rare, very small values are rather rare,
but values close to the middle are rather common. Many
random phenomena obey, at least approximately, a normal
probability distribution. Since this is a good description of lots
of things, the normal distribution is indeed normal!

Examples:
• Height of a person
• Marks obtained by students in a subject
• Error made in measuring a physical quantity
 Normal Distribution (contd.)
A continuous random variable X is normally distributed with
parameters  and 2 i.e., X~N(,2), if its probability density
function (PDF) is given by:
f (x ) =
1 − ( x −  )2 2 2
e −  x  
2 
The normal density f(x) is a bell-shaped curve that is symmetric
about  and which attains a maximum value of 1  2  0.399 
at x= .It can be shown that:
E[X] =  and Var(X) = 2 0.25

The sum of independent normal

0.20

0.15

variables is also a normal random 0.10

variable. 0.05

0.00
25 26 27 28 29 30 31 32 33 34 35

Fish length (in.)

 Normal Distribution (contd.)
Let X~N(μ,σ2) and let Y=aX+b where a and b are constants
Change of scale is the operation of multiplying X by a constant
a because one unit of X becomes “a” units of Y.
Shift is the operation of adding a constant b to X because we
simply move our random variable X “b” units along X axis
If X is a normal random variable, then the new random
variable Y is also a normal random variable with following
parameters
E(Y) =bμ+a
σ2(Y)=b2σ2
Thus, any normal random variable X can be transformed to
another standard normal random variable Y as stated above.
 Standard Normal Distribution
Let X~N(,2)
If we do the transformation b = 1/σ and a = -μ/σ, then
Y = a+bX = (X-μ)/σ
This gives E(Y)=0 and σ2(Y)=1
i.e., Y~N(0,1)
Then Y is said to have standard or unit normal distribution.
Its probability density function (PDF) is given by
f (y) =
1 −( y )2
e 2
−  y  
2
Its cumulative distribution function (CDF) is given as:
(x ) =
1 x
− ( y )2 2

2 
−
e dy −  x  
 Standard Normal Distribution (contd.)
Let X~N(,2) and Y= (X-μ)/σ
Now X<b if (X-μ)/σ < (b-μ)/σ i.e., Y < (b-μ)/σ
 b− b− 
P{ X  b} = P Y   =   
     
Similarly, for any a<b,
a −  b−
P{a  X  b} = P  Y  
   
 b−  a−
= P Y   − P Y  
     
b−  a− 
=   −  
     
Standard Normal Distribution (contd.)

Values of (x) is given in Table A1 for different values of x

(from 0.00 to 3.49).
(x) is 0.5 at x=0 and is 0.9998 at x=3.49
Table A1 tabulates (x) for non-negative values of x.
We can obtain (-x) by making use of symmetry about 0 of
standard normal PDF which means
P{Y<-x} = P{Y>x}
i.e., (-x) = 1-P{Yx}
i.e., (-x) = 1- (x)
 Standard Normal Distribution (contd.)

Problem 1:
If X is random variable with mean 3 and variance 16. Determine
(a) P{X<11} (b) P{X>-1} (c) P{2<X<7}
Transform X to standard normal variable Y=(X-)/
 X − 3 11 − 3  X −3 
(a) PX  11 = P    = P  2
 4 4   4 
=  (2 ) = 0.9772
 X − 3 −1− 3
(b) PX  −1 = P    = PY  −1 = PY  1
 4 4 
=  (1) = 0.8413
 2 − 3 X − 3 7 − 3
(c) P2  X  7 = P     = P− 0.25  Y  1 =  (1) −  (− 0.25)
 4 4 4 
=  (1) − (1 −  (0.25)) = 0.8413 − 1 + 0.5987 = 0.44
 Standard Normal Distribution (contd.)
Problem 2:
Power W dissipated in a resistor =3V2. If V is a normal random
variable with mean 6 and standard deviation 1, determine
(a)E[W] (b) P{W>120}
(a) E[W] = E[3V2] = 3E[V2] = 3(Var[V]+(E[V])2)
= 3(12+62) = 111
Transform V to standard normal variable V=(V-)/
  
(b) PW  120 = P 3V 2  120 = P V  40 
V − 6 40 − 6 
= P   = PY  0.3246
 1 1 
= 1 − PY  0.3246 = 1 −  (0.3246)
= 1 − .6273 = 0.3727
 Exponential Distribution
A continuous random variable X whose probability density
function (PDF) is given for >0, by
f(x) = e-x if x0
=0 if x<0
is known as an exponential random variable with parameter .
Its cumulative distribution F(x) is given by

F ( x ) = PX  x =  e −y dy
x

= 1 − e − x , x0
 Exponential Distribution (Contd.)
• The moment generating function of the exponential is given
by
 (t ) = E e  =  etx e −x dx

tX
0

− ( −t ) x 
= e dx = , t
0  −t
 2
Differentiating,  (t ) = ,  (t ) =
( − t )2
( − t )3
• E[X] = (0) = 1/
• Var(X) = (0)-(E[X])2 = 1/2
 Distributions arising from Normal
•

CHI-SQUARE DISTRIBUTION:
If Z1, Z2,……, Zn are independent standard normal random
variables, then X defined by X = Z12+Z12+…..+Zn2 is said to have
a chi-square distribution with n degrees of freedom. This is
represented as X~n2.
It has additive property.
If X is a chi-square random variable, then for any (0,1), the
quantity 2,n is defined to be such that P{X 2,n}=
Table A2 lists 2,n values for variety of values of  and n.
It is commonly used in statistical significance tests.

•
 Distributions arising from Normal
(contd.)
t-DISTRIBUTION:
If Z and n2 are independent random variables, with Z having
standard normal distribution and n2 having chi-square
distribution with n degrees of freedom, then the random
variable Tn defined by Z
Tn =
 n2 n
is said to have a t-distribution with n degrees of freedom.
t-density is symmetric about zero.
Compared to standard normal density, it has thicker tails
indicating greater variability. As n becomes larger, it becomes
more like standard normal density.
 Distributions arising from Normal
(contd.)
t-DISTRIBUTION (contd.):
Quantity t,n is defined to be such that
P{Tn t,n} = 

It can be shown that

P{Tn -t,n} = 1-

Table A3 lists t,n values for variety of values of  and n

 Sequence of Experiments
We are collecting a large number of data points, and we assume
each data point as the outcome of a random experiment (e.g.
asking for information in a census).

Consider a sequence of n random variables (e.g., throwing a die

n times, doing an experiment n times, or asking for the age of n
residents in a census, etc.). Each outcome is a random variable
Xi, i = 1, 2, 3… n..

1 n
The sample mean X =  Xi is also a random variable itself!!!
n i =1