You are on page 1of 13

INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Course Name: Mathematics & Statistics III (BSC-M301)

1
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Random Variable

Consider the random experiment E of tossing a coin twice in succession.


The sample space S is given by S ={HH, HT, TH,TT} (∵ 22 = 4)
Let us define a variable X as
X – ‘Number of heads obtained’
We see that X(HH) = 2 , X(HT) = 1 = X(TH), X(TT) = 0

Thus we see that X takes the values 0,1 and 2 based on the outcomes of
the sample space.

Hence we observe the following things:


X acts like a function from sample space(S) to the set of real numbers(R)
The domain of X is S while the range of X is {0,1,2}
2
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Such a real valued function X defined over the sample space of any
random experiment is called a random variable.
The word ‘random’ used here refers to the fact that one cannot predict its
value at the first.

There are two types of random variables:

 Discrete random variable

 Continuous random variable

A random variable is denoted by a capital English alphabet


For example: X – ‘Number of heads obtained’
3
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Discrete random variable


A random variable is said to be discrete if its range consists of a finite or at
most countably infinite number of mass/isolated points from the real axis.
For example a random variable X is discrete if it takes up the mass points
𝑥1 , 𝑥2 , 𝑥3 , … … 𝑥𝑛 where n is a finite natural number. Here the range of X is
finite i.e. n
The points 𝑥1 , 𝑥2 , 𝑥3 , … … 𝑥𝑛 are called mass/isolated points.
In the previous example X was a discrete random variable whose range
consist of 3 points and they were 0,1,2.
Since the no. of mass points taken by a discrete random variable is finite or at
most countably infinite so it is possible for us to determine
𝑃 𝑋 = 𝑥1 = 𝑝1 = 𝑓(𝑥1 )
𝑃 𝑋 = 𝑥2 = 𝑝2 = 𝑓(𝑥2 )
⋮ ⋮ ⋮
𝑃 𝑋 = 𝑥𝑛 = 𝑝𝑛 = 𝑓(𝑥𝑛 )
4
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

where 𝑝1 , 𝑝2 , … … 𝑝𝑛 are probabilities that the random variable takes the


values 𝑥1 , 𝑥2 , … … 𝑥𝑛 respectively.
We see that a new function 𝑓 𝑥 is defined which is depended on the mass
points of X and is known as probability mass function(pmf).
For example in the previous example the probability mass function can be
written as
1
𝑃 𝑋=0 = =𝑓 0
4
1
𝑃 𝑋 = 1 = = 𝑓(1)
2
1
𝑃 𝑋 = 2 = = 𝑓(2)
4
It is very easy to observe that for a pmf f(x),
(i) 𝑓 𝑥 ≥ 0
(ii) 𝑥 𝑓 𝑥 = 1 5
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

The distribution of the probability values 𝑝1 , 𝑝2 , … … 𝑝𝑛 over the mass


points 𝑥1 , 𝑥2 , … … 𝑥𝑛 of a discrete random variable X can be written in a
tabular form as:

𝑃(𝑋 = 𝑥𝑖 ) 𝑥1 𝑥2 ………… 𝑥𝑛 Total

𝑓(𝑥𝑖 ) 𝑝1 𝑝2 …………. 𝑝𝑛 1

Such a distribution of the probabilities over the mass points of a discrete


random variable X is known as discrete probability distribution.

6
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Thus the discrete probability distribution of the random variable X of the


previous example is given by

𝑃(𝑋 = 𝑥𝑖 ) 0 1 2 Total

𝑓(𝑥𝑖 ) 1 1 1 1
4 2 4
Clearly we see that 𝑓 𝑥 ≥ 0
And also we find that,
1 1 1
𝑓 𝑥 = + + =1
4 2 4
𝑥
Hence f(x) is a possible probability mass function(pmf)

7
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Continuous Random variable

A random variable is said to be continuous if the range of X is an interval


i.e. if X is defined within an interval
For example (say) a < X < b or 𝑎 ≤ 𝑋 < 𝑏 𝑜𝑟 𝑋 ≥ 𝑎 𝑜𝑟 𝑋 ≤ 𝑏 𝑒𝑡𝑐.
Thus the number of mass points taken by a continuous random variable X
is uncountably infinite.
Let us consider a continuous random variable X defined as a < X < band let
a < c < b.
Now to find P (X = c ) is impossible since the number of mass points for a
cont. random variable being uncountably infinite.
Hence as a result we assume P( X = c ) = 0
Thus for each and every mass points in the interval (a , b) the probability
that X takes a specific value is assumed to be zero.
8
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Thus, the question is whether a probability function corresponding to


continuous random variable can be defined or not?
If we consider a continuous random variable X defined in the interval (a , b)
𝑑𝑥 𝑑𝑥
and let (𝑥 − , 𝑥 + ) be an infinitesimal small sub interval of (a, b) then
2 2
we see that
𝑑𝑥 𝑑𝑥
P (𝑥 − < X <𝑥 + ) >0
2 2
This is so because if we use the concept of definite integral as a limit of sum
then we will get
𝑑𝑥
𝑑𝑥 𝑑𝑥 𝑥+ 2
P (𝑥 − < X <𝑥 + ) = 𝑑𝑥 𝑓 𝑥 𝑑𝑥 > 0 , the function f(x) is known as
2 2 𝑥− 2
probability density function (pdf)
It is easy to see that if dx = 0 then P ( X = x) = 0, where x is a specific point in
the interval (a , b) as we have noticed earlier.
9
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Continuous Random Variable (Contd.)

Thus we are able to define a probability function corresponding to a


continuous random variable and is known as probability density
function(pdf).
We observe two important properties of a pdf and they are
 𝑓(𝑥) ≥ 0


 −∞
𝑓 𝑥 𝑑𝑥 = 1
Thus for any continuous random variable X the probability that it lies in an
𝑏
interval (a , b) is given as 𝑃 𝑎 ≤ 𝑋 ≤ 𝑏 = 𝑎 𝑓 𝑥 𝑑𝑥.
This actually refers to the bounded area under the curve f(x), x = a , x = b
and x axis.
10
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Geometrically 𝑃(𝑎 ≤ 𝑋 ≤ 𝑏) represents the shaded area in the figure

But the question is whether 𝑃 𝑎 < 𝑋 < 𝑏 𝑎𝑛𝑑 𝑃(𝑎 ≤ 𝑋 ≤ 𝑏) are equal or
not?

11
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

To prove this, we consider a continuous random variable X in [a , b]


Then we can write
𝑎 ≤𝑋 ≤𝑏 =𝑎 <𝑋 <𝑏+𝑋 =𝑎+𝑋 =𝑏
Now taking probability on both sides we get,
𝑃 𝑎 ≤𝑋 ≤𝑏 =𝑃 𝑎 <𝑋 <𝑏 +𝑃 𝑋 =𝑎 +𝑃 𝑋 =𝑏
(since the events 𝑎 < 𝑋 < 𝑏, X = a and X = b are pairwise mutually exclusive)
As X is a continuous random variable so P(X= a) = P( X = b) = 0
∴ 𝑃 𝑎 ≤ 𝑋 ≤ 𝑏 = 𝑃(𝑎 < 𝑋 < 𝑏)
Similarly we can show that 𝑃 𝑎 ≤ 𝑋 < 𝑏 = 𝑃(𝑎 < 𝑋 < 𝑏)&
𝑃 𝑎 < 𝑋 ≤ 𝑏 = 𝑃(𝑎 < 𝑋 < 𝑏)

Thus we have the following result


𝑏
𝑃 𝑎 ≤𝑋 ≤𝑏 =𝑃 𝑎 <𝑋 <𝑏 =𝑃 𝑎 ≤𝑋 <𝑏 =𝑃 𝑎 <𝑋 ≤𝑏 = 𝑎
𝑓 𝑥 𝑑𝑥
12
INSTITUTE OF ENGINEERING & MANAGEMENT, KOLKATA

Thank You

13

You might also like