Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Lecture 4: Continuous Random Variables and Probability

Distributions

Faculty of Mathematics and Computer Science, University of Science

Vietnam National Uuniversity Ho Chi Minh city

Degree A. Y. 2024 - 2025

Advanced Program in Computer Science
Course: Applied Statistics for Engineers and Scientists I

1 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Outline

1 Continuous Random Variables

2 Expected value and Variance

3 Some continuous distributions

4 The approximate distribution

2 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

1 Continuous Random Variables

2 Expected value and Variance

3 Some continuous distributions

4 The approximate distribution

3 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Continuous Random Variables

Continuous Random Variables

A random variable X is continuous if its value space SX is an interval of real
numbers, either of the form (a, b) or the entire real number space R.

Since the possible values of X are infinite, we have:

P(X = x ) = 0;
a probability distribution table cannot be used to describe X .

Instead, X is represented as:

a number line over R, along with
a function f (x ) describing the density of values of X over a given interval.

4 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Continuous Random Variables

Example 1: Consider a continuous random variable X with a visible value

range of (−4, 4). The density of values x is represented by the function f (x ) as
shown in the figure.

0.4

0.3
f(x)

0.2

0.1

0.0
−4 −2 0 2 4
x

The larger f (x ) is, the denser the values around x are.

5 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Probability Density Function

robability Density Function

The function that describes the density of a continuous random variable X over
the value space SX , is called the probability density function (pdf), denoted
as f (x ), with the following properties:
f (x ) is integrable;
f (x ) ≥ 0, for all values of x ;
Z
f (x ) dx = 1;
S
Z b
P(a ≤ X ≤ b) = f (x ) dx , which is the area under the curve of f (x )
a
over the interval [a, b] of X .

The probability density function f (x ) is explicitly defined by a mathematical

formula.

6 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Probability Density Function

Example 2: Consider a continuous random variable X with a value space SX
on the interval [−1, 1], has the probability density function:

0.5 if − 1 ≤ x ≤ 1;
f (x ) =
0 if x < −1 or x > 1.
1.00

0.75
f(x)

0.50

0.25

0.00
−1 0 1
x

7 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Probability Density Function

Example 3: Consider a continuous random variable X with a value space

SX ≡ R, has the probability density function:
 
1 x2
f (x ) = √ exp −
2π 2

0.5

0.4

0.3
f(x)

0.2

0.1

0.0
−4 −2 0 2 4
x

8 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Cumulative Probability Function

Cumulative Distribution Function

Consider a continuous random variable X with value space R and pdf f (x ).
The cumulative distribution function (cdf) of X is defined as:
Zx
F (x ) = P(X ≤ x ) = f (s) ds.
−∞

The cdf F (x ) has the following properties:

F (x ) is continuous and differentiable;
dF (x )
the derivative of F (x ) is f (x ), i.e., = f (x ).
dx

9 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Cumulative Probability Function

Example 2: (continued). Consider a continuous random variable X with a

value space SX on the interval [−1, 1], has the probability density function:

0.5 if − 1 ≤ x ≤ 1;
f (x ) =
0 if x < −1 or x > 1.

1.00

We calculate: 0.75

0 if x < −1
F(x)

0.50
x +1
F (x ) = if − 1 ≤ x ≤ 1;
 2

1 if x > 1. 0.25

0.00
−1 0 1
x

10 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Cumulative Probability Function

Example 3: (continued). Consider a continuous random variable X with a

value space SX ≡ R, has the probability density function:
 
1 x2
f (x ) = √ exp −
2π 2

1.00

0.75
We have
Zx
F(x)
 
1 s2 0.50
F (x ) = √ exp − ds
2π 2
−∞ 0.25

0.00
−4 −2 0 2 4
x

11 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

The probability P(X ≤ a) is the area under the pdf f (x ):

0.5

0.4

0.3
f(x)

0.2
P(X ≤ a)

0.1

0.0
−4 −2 a 0 2 4

and computed as
Za
P(X ≤ a) = f (x ) dx = F (a)
−∞

12 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

The probability P(X ≥ a) is the area under the pdf f (x )

0.5

0.4

0.3 P(X ≥ a)
f(x)

0.2

0.1

0.0
−4 −2 a 0 2 4
x

and computed as
Z∞
P(X ≥ a) = f (x ) dx = 1 − P(X ≤ a) = 1 − F (a)
a

13 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

The probability P(a ≤ X ≤ b) is the area under the pdf f (x )

0.5

0.4

0.3 P(a ≤ X ≤ b)
f(x)

0.2

0.1

0.0
−4 −2 a 0 b 2 4
x

and computed as
Zb
P(a ≤ X ≤ b) = f (x ) dx = P(X ≤ b) − P(X ≤ a) = F (b) − F (a)
a

14 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

Note: Due to the continuity of the cdf F (x ), we have:

P(X < a) = P(X ≤ a),

similarly,
P(a < X < b) = P(a ≤ X ≤ b).
Exercise 1: Let X be a continuous random variable representing the measured
current intensity in a thin copper wire, in milliamperes. Assume that the range
of X is [0, 20mA] and that its probability density function is:

0.05 if 0 ≤ x ≤ 20;
f (x ) =
0 if x < 0 or x > 20.

(a) Determine the cumulative distribution function F (x ).

(b) Compute P(X ≤ 10).
(c) Compute P(X > 15).
(d) Compute P(5 ≤ X ≤ 18).
Answers: (b) 0.5; (c) 0.25; (d) 0.65
15 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

Exercise 2: Consider a continuous random variable X with pdf as f (x ) = e −x
if x > 0, otherwise f (x ) = 0. Determine:

(a) the cdf F (x ), (f) P(3 ≤ X ),

(b) P(1 < X ), (g) the value of x such that
(c) P(1 < X < 2.5), P(x < X ) = 0.1,
(d) P(X = 3), (h) the value of x such that
(e) P(X < 4), P(X ≤ x ) = 0.1.

Answers: (b) 0.368; (c) 0.286; (d) 0; (e) 0.982; (f) 0.05; (g) 2.303; (h) 0.095
Exercise 3: Consider a continuous random variable X with pdf as
f (x ) = 0.5 cos(x ) if −π/2 < x < π/2, otherwise f (x ) = 0. Determine:

(a) cdf F (x ), (e) P(X > −π/4),

(b) P(X < 0),
(c) P(X < −π/4), (f) the value of x such that
(d) P(−π/4 < X < π/4), P(X < x ) = 0.95.

Answers: (b) 0.5; (c) 0.146; (d) 0.707; (d) 0.854; (f) 1.120
16 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of continuous RVs

Exercise 4: Consider the density curve shown in the figure, representing the
distribution of diameters (measured 4.5 feet above the ground) in a population
of 30-year-old Douglas fir trees. The areas under the curve are shown in the
figure. What is the probability that a tree has a diameter:
(a) between 4 inches and 10 inches?
(b) less than 4 inches?
(c) more than 6 inches?

Answers: (a) 0.7; (b) 0.23; (c) 0.44

17 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

1 Continuous Random Variables

2 Expected value and Variance

3 Some continuous distributions

4 The approximate distribution

18 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Expected value and Variance of continuous RVs

Just like discrete random variables, for continuous random variables, we
consider two important characteristics:
the expected value (mean)
variance.
Expected value and Variance of continuous RVs
Consider X as a continuous random variable on R, with probability density
function f (x ). Then, the expected value (or mean) of X , denoted as E(X ) or
µX , is given by:
Z+∞
E(X ) = x f (x ) dx
−∞

The variance of X denoted as Var(X ) or σX2 , is given by:

Z+∞ Z+∞
2
Var(X ) = (x − E(X )) f (x ) dx = x 2 f (x ) dx − E(X )2 .
−∞ −∞
p
The square root of the variance, Var(x ) is called the standard deviation of
X , denoted as σX .
19 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Expected value and Variance of continuous RVs

Exercise 5: Let X be a continuous random variable with the probability density

function: 
0.05 if 0 ≤ x ≤ 20;
f (x ) =
0 if x < 0 or x > 20.
Find the mean and standard deviation of X .
Answer: E(X ) = 10; Var(X ) = 33.333, σ = 5.774
Exercise 6: Consider a continuous random variable X with pdf as f (x ) = e −x
if x > 0, otherwise f (x ) = 0. Find the mean and standard deviation of X .
Answer: E(X ) = 1; Var(X ) = 1, σX = 1
Exercise 7: Consider a continuous random variable X with pdf as
f (x ) = 0.5 cos(x ) if −π/2 < x < π/2, otherwise f (x ) = 0. Find the mean and
standard deviation of X .
Answer: E(X ) = 0; Var(X ) = 0.467, σX = 0.683

20 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Some Properties

Here are some important properties of the expected value and variance that
should be remembered:
E(c) = c, for c is a constant.
E(cX + b) = cE(X ) + b, for c and b are constants.
Given two RVs X and Y , we have

E(aX + bY ) = aE(X ) + bE(Y ).

Given two RVs X and Y , if X and Y are independent, then

E(XY ) = E(X )E(Y ).

p
Var(X ) ≥ 0 and σ = Var(X ) ≥ 0.
Var(X ) = E (X − E(X ))2 = E X 2 − E(X )2 .

21 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Some Properties

Var(c) = 0, for c is a constant.

Var(cX + b) = c 2 Var(X ), for c and b are constants.
Given a function h(X ) of X , then

E(h(X )) = h(x1 )f (x1 ) + h(x2 )f (x2 ) + . . . + h(xn )f (xn ),

and
Var(h(X )) = E(h(X )2 ) − E(h(X ))2 .
Given two RVs X and Y , if X and Y are independent, then

Var(aX + bY ) = a2 Var(X ) + b 2 Var(Y ).

If X has a unit (e.g., m, s, kg, VND, etc.), then:

the expected value and standard deviation of X will have the same unit as
X;
the variance will have the square of the original unit (e.g., m2 , s2 , kg2 ,
VND2 , etc.).

22 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

1 Continuous Random Variables

2 Expected value and Variance

3 Some continuous distributions

4 The approximate distribution

23 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Exponential distribution

Exponential distribution
A continuous random variable X is
called to follow an exponential 2.0

distribution if its probability density

λ
function is given by: 1.5
2
1

f(x)
−λx
f (x ) = λe , 1.0
0.5

for x ≥ 0, and parameter λ > 0. 0.5

The cumulative distribution function of 0.0

0 1 2 3 4 5 6 7 8 9
X is: x
F (x ) = 1 − e −λx ,
for x ≥ 0.
The mean and variance of X are given by
1 1
E(X ) = , Var(X ) = .
λ λ2
We denote this distribution as X ∼ Exp(λ).
24 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Exponential distribution

Exercise 8: Consider a random variable X that follows an exponential

distribution with parameter λ = 2. Determine:

(a) the cdf F (x ), (e) P(1 < X < 2),

(b) P(X ≤ 0),
(c) P(X ≥ 2), (f) the value of x such that
(d) P(X ≤ 1), P(X < x ) = 0.05.

Answers: (b) 0; (c) 0.0183; (d) 0.865; (e) 0.117; (f) 0.026
Exercise 9: Consider a random variable X that follows an exponential
distribution with a mean of 10. Determine:

(a) the cdf F (x ), (e) P(15 < X < 25),

(b) P(X > 10),
(c) P(X > 20), (f) the value of x such that
(d) P(X < 30), P(X < x ) = 0.90.

Answers: (b) 0.368; (c) 0.135; (d) 0.950; (e) 0.141; (f) 23.026

25 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Exponential distribution and Applications

The exponential distribution is often used to describe the time between

occurrences of events:

Let Y represent the number of times an event occurs within a time interval t,
then,
Y ∼ P(tλ),
i.e., Y follows a Poisson distribution with parameter tλ.

26 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Exponential distribution and Applications

Example 4: A home appliance company records that the time between two
consecutive warranty claims for their products follows an exponential
distribution with a mean of 1.5 days (1 day = 24 hours).
Let X be the “time between two consecutive warranty claims.” Therefore,
X ∼ Exp(λ).
1 1
According to the description, we have E(X ) = = , which implies
λ 1.5
λ = 0.667
Let Y be the “number of warranty claims received within 5 hours”. Then,
5
 
Y ∼ P 0.667 ×
24
or Y ∼ P(0.139).
In this example, the time frame T is in days, while t is 5 hours. Therefore, we
need to convert t into days.

27 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Exponential distribution and Applications

Exercise 10: A home appliance company records that the time between two
consecutive warranty claims for their products follows an exponential
distribution with a mean of 1.5 days (1 day = 24 hours).
(a) Calculate the probability that no warranty claims are received by the
company within 5 hours.
(b) Calculate the probability that the company receives at least 2 warranty
claims within 5 hours.
(c) A random day is selected for observation. Given that the company
received more than 1 warranty claim on that day, calculate the probability
that no more than 3 warranty claims were received on that day.

Answers: (a) 0.8703; (b) 0.0088; (c) 0.9664

28 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Distribution

A continuous random variable X follows a normal distribution if its probability

density function is given by:
 
1 (x − µ)2
f (x ) = √ exp − ,
σ 2π 2σ 2

where −∞ < x < +∞, and the two parameters −∞ < µ < +∞, and σ > 0.
For example, given µ = 0 and σ 2 = 1, the pdf is shown as follows:

0.5
µ=0
0.4 σ2 = 1

0.3
f(x)

0.2

0.1

0.0
−4 −2 0 2 4
x

29 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Distribution

The cumulative distribution function of X is:

Zx  
1 (s − µ)2
F (x ) = √ exp − ds.
σ 2π 2σ 2
−∞

The mean and variance are given by:

E(X ) = µ, Var(X ) = σ 2

We denote this as X ∼ N(µ, σ 2 ), meaning X follows a normal distribution with

mean µ and variance σ 2 .

30 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Distribution

Changing the values of µ and σ 2 results in different shapes of f (x ):

keeping σ fixed, changing µ shifts the position of f (x );
keeping µ fixed, increasing or decreasing σ, change the spread (width) of
f (x );
The shape of f (x ) is always (bell-shaped) and symmetric about µ,

31 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Distribution

For X ∼ N(µ, σ 2 ), the following probabilities hold:

P(µ − σ ≤ X ≤ µ + σ) = 0.6827,
P(µ − 2σ ≤ X ≤ µ + 2σ) = 0.9545,
P(µ − 3σ ≤ X ≤ µ + 3σ) = 0.9973.
These results are known as the “kσ rule”.

32 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Standard Normal Distribution

Standard Normal Distribution - Gauss

For random variable Z ∼ N(µ, σ 2 ), if µ = 0 and σ = 1, we say Z follow a
standard normal distribution (or Gauss distribution), denoted as Z ∼ N(0, 1).
The cdf of Z is denoted as Φ(z):
Zz  
1 s2
Φ(z) = P(Z ≤ z) = √ exp − ds.
2π 2
−∞
Values of Φ(z) are summarized in the Table A.3 of textbook.

0.5
P(X ≤ 1.04) = Φ(1.04) z 0.00 0.01 0.02 0.03 0.04
0.4
0.0 0.5000 0.5040 0.5080 0.5120 0.5160
0.3 0.1 0.5398 0.5438 0.5478 0.5517 0.5557
f(x)

. . . . . .
. . . . . .
0.2 . . . . . .
1.0 0.8413 0.8438 0.8461 0.8485 0.8508
0.1 . . . . . .
. . . . . .
. . . . . .
0.0
−4 −2 0 1.04 2 4

33 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Standard Normal Distribution

For Z ∼ N(0, 1), we have:

Φ(a) = 1 − Φ(−a), for any value of a,
P(Z > a) = 1 − Φ(a),
P(a ≤ X ≤ b) = Φ(b) − Φ(a),

Example 5: Using Table A.3, calculate:

(a) P(Z > 1.26) (c) P(Z > −1.37)

(b) P(Z < −0.86) (d) P(−1.25 < Z < 0.37)

Solutions:
(a) P(Z > 1.26) = 1 − P(Z ≤ 1.26) = 1 − Φ(1.26) = 1 − 0.8962 = 0.1038;
(b) P(Z < −0.86) = 1 − Φ(0.86) = 1 − 0.8051 = 0.1949;
(c) P(Z > −1.37) = 1 − P(Z < −1.37) = 1 − (1 − Φ(1.37)) = Φ(1.37) =
0.9147;
(d) P(−1.25 < Z < 0.37) = Φ(0.37) − Φ(−1.25) = Φ(0.37) − 1 + Φ(1.25) =
0.5387

34 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Standard Normal Distribution

Exercise 11: Using Table A.3, calculate:

(a) P(Z < 1.32), (d) P(Z > −2.15),

(b) P(Z < 3.0),
(c) P(Z > 1.45), (e) P(−2.34 < Z < 1.76).

Exercise 12: Using Table A.3, calculate:

(a) P(−1 < Z < 1), (d) P(Z > 3),

(b) P(−2 < Z < 2),
(c) P(−3 < Z < 3), (e) P(0 < Z < 1).

Exercise 13: Using Table A.3, find the value of z such that:

(a) P(Z < z) = 0.9, (d) P(Z > z) = 0.9,

(b) P(Z < z) = 0.5,
(c) P(Z > z) = 0.1, (e) P(−1.24 < Z < z) = 0.8.

35 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of normal RVs

Standardized normal RVs

Consider X ∼ N(µ, σ 2 ), the transformation
X −µ
Z =
σ
resulting Z ∼ N(0, 1). This transformation is called “standardized normal
transformation”.

This transformation allows us to calculate the probability of X through Z . In

particular:
X −µ x −µ x −µ x −µ
     
P(X ≤ x ) = P ≤ =P Z ≤ =Φ .
σ σ σ σ
a−µ X −µ b−µ
 
P(a ≤ X ≤ b) = P ≤ ≤ =
σ σ σ
b−µ a−µ
   
Φ −Φ ,
σ σ
x −µ
 
P(X ≥ x ) = 1 − P(X ≤ x ) = 1 − Φ ,
σ
a−µ a−µ
   
P(X ≤ a) = Φ =1−Φ − .
σ σ
36 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of normal RVs

Example 6: Given X ∼ N(10, 4), calculate
(a) P(X ≤ 12),
(b) P(9 < X < 11),
(c) find the value of x such that P(X < x ) = 0.98

X − 10 12 − 10 12 − 10
   
(a) P(X ≤ 12) = P ≤ =P Z ≤ = Φ(1) =
2 2 2
0.8413.
11 − 10 9 − 11
   
(b) P(9 < X < 11) = Φ −Φ = Φ(0.5) − Φ(−0.5) =
2 2
Φ(0.5) − 1 + Φ(0.5) = 0.3830.
(c) Using standardized normal transformation, we have
x − 10
 
P(X < x ) = Φ = 0.98
2
By using Table A.3 Φ(z), we obtain Φ(2.05) = 0.9798 ≈ 0.98, therefore
x − 10
= 2.05
2
It implies x = 2 × 2.05 + 10 = 14.1
37 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of normal RVs

Exercise 14: Let X be a random variable that follows a normal distribution

with mean 5 and standard deviation 4. Calculate the following probabilities:

(a) P(X < 11), (d) P(−2 < X < 9),

(b) P(X > 0), (e) P(2 < X < 8),
(c) P(3 < X < 7), (f) P(1 ≤ X ≤ 6).

Exercise 15: Let X be a random variable that follows a normal distribution

with mean 10 and standard deviation 2. Find the value of x such that:

(a) P(X > x ) = 0.5, (d) P(−4 < X − 10 < x ) = 0.95,

(b) P(X > x ) = 0.95,
(c) P(x < X < 10) = 0.2, (e) P(−x < X − 10 < x ) = 0.99.

38 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Compute the probability of normal RVs

Exercise 16: (Laptop Battery Life)The time until a laptop battery needs
recharging follows a normal distribution with mean 260 minutes and standard
deviation 50 minutes.
(a) What is the probability that a battery lasts more than 4 hours?
(b) Find the 1st quartile and 3rd quartile of battery life. (x is the 1st quartile
if P(X ≤ x ) = 0.25; x x is the 3rd quartile if P(X ≤ x ) = 0.75.)
(c) What battery life (in minutes) has a greater than 95% probability?
Answers: (a) 0.6554; (b) 226.276 và 293.725; (c) 342.243

Exercise 17: (Sunflower Growth) The 15-day growth of a sunflower population

follows a normal distribution with mean 3.18 cm standard deviation 0.53 cm.
(a) What percentage of sunflowers grow 4 cm or more?
(b) What percentage grow 3 cm or less?
(c) What percentage grow between 2.5 cm and 3.5 cm?
Answers: (a) 0.061; (b) 0.367; (c) 0.627

39 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Uniform distribution

A uniform distribution, sometimes also known as a rectangular

distribution, is a distribution that has constant probability over a given
range (a, b) with a < b.
The uniform distribution is denoted as U(a, b), with a and b are the lower
bound and upper bound of the range.
a+b (b − a)2
The mean µ = (also for median) and the variance σ 2 = ,
2 12
the mode is any value in the range (a, b).
The density function f (x ) is defined as
( 1
if a ≤ x ≤ b
f (x ) = b−a
0 if x > b or x < a

40 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Uniform distribution

41 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

t-distribution

The t-distribution is a bell-shaped distribution, similar to the normal

distribution, but with heavier tails.
The mean, median, and mode of the t-distribution are equal to 0.
The t-distribution is a family of curves, each determined by a parameter
called the degrees of freedom - df .
The t-distribution is denoted as tn , with n as the degree of freedom.
As the degrees of freedom increase, the t-distribution approaches the
standard normal distribution. Often, after 30 df, the t-distribution is very
close to the standard normal distribution N (0, 1).
The area under each curve of the t-distribution equals 1.

42 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

t-distribution

0.4
0.3
Standard normal distribution
0.2
0.1

t−distribution, df = 5 t−distribution, df = 1
0.0

−4 −2 0 2 4

43 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Chi-square χ2 distribution

The distribution for a random variable that has positive value (greater
than or equal to 0).
The Chi-square distributions are right-skewed.
The Chi-square distribution is a family of curves, each determined by the
degrees of freedom.
The Chi-square distribution is denoted by χ2n , with n as degree of freedom.
The area under each curve of the chi-square distribution equals 1.
Assumption that we have k random variable X1 , X2 , . . . , Xk , such as
Xi ∼ N (0, 1) with i = 1, 2, . . . , k. Then, Xi2 ∼ χ21 and
Y = X12 + X22 + . . . + Xk2 ∼ χ2k .

44 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Chi-square χ2 distribution

0.5
0.4
0.3

Chi−square distribution, df = 2
0.2

Chi−square distribution, df = 5

Chi−square distribution, df = 10
0.1
0.0

0 5 10 15 20

45 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

1 Continuous Random Variables

2 Expected value and Variance

3 Some continuous distributions

4 The approximate distribution

46 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

Normal Approximation of Binomial Distribution

Consider a sequence of independent and identically distributed (i.i.d.) Bernoulli
random variables, denoted as X1 , X2 , . . . , Xn , where Xi ∼ B(p) with E(Xi ) = p
and Var(Xi ) = p(1 − p), for all i = 1, 2, . . . , n.
Let
n
X
X = Xi
i=1

then X ∼ B(n, p). When n is sufficiently large, specifically when np ≥ 5 and

n(1 − p) ≥ 5, we can approximate
X − np
p ∼ N(0, 1),
np(1 − p)

47 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

48 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

Based on this result, we can use the normal distribution to approximate the
probability of a binomial variable. Specifically:
!
X − np x + 0.5 − np
P(X ≤ x ) = P(X ≤ x + 0.5) = P p ≤ p
np(1 − p) np(1 − p)
!
x + 0.5 − np
≈ Φ p
np(1 − p)

and
!
X − np x − 0.5 − np
P(X < x ) = P(X ≤ x − 0.5) = P p ≤ p
np(1 − p) np(1 − p)
!
x − 0.5 − np
≈ Φ p
np(1 − p)

where Φ(z) is the cumulative distribution function (CDF) of the standard

normal distribution N(0, 1).

49 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

Example 7: A communication channel has a bit error probability of p = 0.1.

Suppose we transmit 50 bits of information. Let X be the total number of
erroneous bits, then X ∼ B(50, 0.1).
(a) Computing P(X ≤ 2). Using the binomial probability formula:
0
P(X ≤ 2) = C50 × 0.10 × 0.950 + C50
1
× 0.11 × 0.949 + C50
2
× 0.12 × 0.948
= 0.112

Using normal approximation with 50 × 0.1 = 5 and 50 × 0.9 = 45:

P(X ≤ 2) = P(X ≤ 2 + 0.5)

 
2 + 0.5 − 50 × 0.1
≈ Φ √
50 × 0.1 × 0.9
= Φ(−1.18) = 0.119

50 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

(b) Computing P(X > 8):

8
X i
P(X > 8) = 1 − P(X ≤ 8) = 1 − C50 × 0.1i × 0.950−i
i=0

Instead of computing directly, using normal approximation:

P(X > 8) = 1 − P(X ≤ 8) = 1 − P(X ≤ 8 + 0.5)

 
8 + 0.5 − 50 × 0.1
≈ 1−Φ √
50 × 0.1 × 0.9
= 1 − Φ(1.65) = 0.05

The results show that when n is large enough, using normal approximation
simplifies probability calculations compared to the exact binomial probability
formula.

51 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Binomial Distribution

Exercise 18: Given X ∼ B(200, 0.4), compute:

(a) P(X ≤ 70),
(b) P(70 ≤ X ≤ 90),
(c) P(X = 80).
Answers: (a) 0.085; (b) 0.87; (c) 0.057

Exercise 19: A semiconductor manufacturing process produces 2% defective

chips. Assume the chips are independent, and each batch contains 1000 chips.
(a) Estimate the probability that more than 25 chips are defective.
(b) Estimate the probability that between 20 and 30 chips are defective.
Answers: (a) 0.107; (b) 0.536

52 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Poisson Distribution

Normal Approximation of Poisson Distribution

Given a sequence of independent and identically distributed Poisson random
variables X1 , X2 , . . . , Xn with mean λ, i.e., Xi ∼ P(λ) with E(Xi ) = λ and
Var(Xi ) = λ, for all i = 1, 2, . . . , n.
Let
n
X
X = Xi ,
i=1

then X ∼ P(nλ). When n is sufficiently large, we can approximate

X − nλ
√ ∼ N(0, 1),
nλ
In particular, for n = 1, we have X ∼ P(λ) and
X −λ
√ ∼ N(0, 1),
λ
provided that λ ≥ 5.

53 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Poisson Distribution

Based on this result, we can use the normal distribution to approximate the
probability of a Poisson variable. Specifically:
 
X −λ x + 0.5 − λ
P(X ≤ x ) = P(X ≤ x + 0.5) = P √ ≤ √
λ λ
 
x + 0.5 − λ
≈ Φ √
λ
and
 
X −λ x − 0.5 − λ
P(X < x ) = P(X ≤ x − 0.5) = P √ ≤ √
λ λ
 
x − 0.5 − λ
≈ Φ √
λ

54 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Poisson Distribution

Example 8: Suppose the number of asbestos fibers in one square meter of dust
on a surface follows a Poisson distribution with an average of 1000. If we
analyze one square meter of dust, what is the probability of finding 950 or
fewer fibers?
Solution: We have X ∼ P(1000). The probability to compute is P(X ≤ 950).
Using the Poisson probability formula, we get:
950
X 1000x e −1000
P(X ≤ 950) =
x!
x =0

This formula is computationally cumbersome, even when considering its

complement.
Instead, since λ = 1000 is very large, so we approximate:

P(X ≤ 950) = P(X ≤ 950 + 0.5)

 
950 + 0.5 − 1000
≈ Φ √
1000
= Φ(−1.57) = 0.058
55 / 56
Continuous Random Variables Expected value and Variance Some continuous distributions The approximate distribution

Normal Approximation of Poisson Distribution

Exercise 20: Suppose the random variable X follows a Poisson distribution

with λ = 6.
(a) Using the Poisson distribution, calculate P(X ≤ 4).
(b) Using the normal approximation, compute P(X ≤ 4) and compare it with
part (a).
(c) Approximate the probability P(8 < X < 12).
Answers: (a) 0.285; (b) 0.270; (c) 0.141

Exercise 21: A high-volume printer produces minor print quality defects on a

sample of 1000 pages according to a Poisson distribution with an average of
0.4 defects per page.
(a) Why are the number of defects per page independent random variables?
(b) What is the expected number of pages with at least one defect?
(c) Estimate the probability that more than 350 pages have at least one
defect.
Answers: (b) 329.68; (c) 0.081

56 / 56