SEHH2311

FOUNDATIONS OF DATA SCIENCE

LECTURE 1
Revision of Essentials of Probability and
Statistics
 Topics
1. Discrete Random Variables
2. Continuous Random Variables
3. Expectations of Random Variables
4. Variances of Random Variables
5. Distributions of Functions of Random Variables

SEHH2311 Foundations of Data Science Page 2

 Random Variable
• A random variable is a variable whose value is
determined by random events
• Random variable can be discrete or continuous
• Example
– Height of a randomly selected students from HKCC is a
continuous random variable (X)
– Number of family members in a randomly selected
household from an estate is a discrete random variable (Y)
• We typically use a capital letter to denote a random
variable and small letters to denote the "observed"
data values of a random variable
SEHH2311 Foundations of Data Science Page 3
 Discrete Random Variable
• The behavior of a discrete random variable can be fully
described by using a probability mass function (p.m.f.)
• Probability Mass Function 𝒇𝒇(𝒙𝒙) of a discrete random
variable 𝑋𝑋 is the probability 𝑃𝑃(𝑋𝑋 = 𝑥𝑥)
• Sample space 𝑺𝑺 is the set of possible values of 𝑋𝑋
• An event regarding the random variable 𝑋𝑋 corresponds
to a subset of values of 𝑆𝑆, for example, 𝐴𝐴 ⊂ 𝑆𝑆

SEHH2311 Foundations of Data Science Page 4

 Discrete Random Variable
Properties of p.m.f.
Properties of a pmf 𝑓𝑓(𝑥𝑥)
1. 𝑓𝑓 𝑥𝑥 > 0, 𝑥𝑥 ∈ 𝑆𝑆
2. ∑𝑥𝑥∈𝑆𝑆 𝑓𝑓(𝑥𝑥) = 1
3. 𝑃𝑃 𝑋𝑋 ∈ 𝐴𝐴 = ∑𝑥𝑥∈𝐴𝐴 𝑓𝑓 𝑥𝑥 , where 𝐴𝐴 ⊂ 𝑆𝑆

SEHH2311 Foundations of Data Science Page 5

 Example 1
Let 𝑋𝑋 be the number showing up when rolling a die.

Sample space: 𝑆𝑆 = 1, 2, 3, 4, 5, 6
pmf: 𝑓𝑓 𝑥𝑥 = 1/6 for 𝑥𝑥 ∈ 𝑆𝑆
i.e. 𝑓𝑓 1 = 𝑓𝑓 2 = ⋯ = 𝑓𝑓 6 = 1/6

Suppose we are interested in the event that the number showing up

is even. Then the set 𝐴𝐴 is 2, 4, 6
𝑃𝑃 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = � 𝑓𝑓(𝑥𝑥) = 𝑓𝑓 2 + 𝑓𝑓 4 + 𝑓𝑓 6 = 1/2
𝑥𝑥∈𝐴𝐴

SEHH2311 Foundations of Data Science Page 6

 Example 2
Suppose we have two $2 coins, three $5 coins and five $10 coins in
a box. Let 𝑋𝑋 be the "value" of a coin drawn randomly from the box.

Sample space: 𝑆𝑆 = {$2, $5, $10}

2 3 5
pmf: 𝑓𝑓 2 = = 0.2, 𝑓𝑓 5 = = 0.3, 𝑓𝑓 10 = = 0.5
10 10 10

Suppose we are interested in the event of drawing a coin of value

greater than $2. i.e. 𝐴𝐴 = $5, $10
𝑃𝑃 𝑋𝑋 > 2 = � 𝑓𝑓(𝑥𝑥) = 𝑓𝑓 5 + 𝑓𝑓 10 = 0.8
𝑥𝑥∈𝐴𝐴

SEHH2311 Foundations of Data Science Page 7

 Example 3
Suppose 𝑋𝑋 is the number of Heads showing up when
flipping a fair coin 3 times. Find the sample space 𝑆𝑆 and
the probability mass function of 𝑋𝑋.
Solution:

SEHH2311 Foundations of Data Science Page 8

 Example 3
Suppose 𝑋𝑋 is the number of Heads showing up when
flipping a fair coin 3 times. Find the sample space 𝑆𝑆 and
the probability mass function of 𝑋𝑋.
Solution:
Sample space: 𝑆𝑆 = {0, 1, 2, 3}

Probability Mass Function:

𝑓𝑓 0 = 𝑃𝑃 𝑋𝑋 =0 = 1/8
𝑓𝑓 1 = 𝑃𝑃 𝑋𝑋 =1 = 3/8
𝑓𝑓 2 = 𝑃𝑃 𝑋𝑋 =2 = 3/8
𝑓𝑓 3 = 𝑃𝑃 𝑋𝑋 =3 = 1/8

SEHH2311 Foundations of Data Science Page 9

 Discrete Random Variable
Cumulative Distribution Function (CDF)
Cumulative Distribution Function 𝑭𝑭(𝑿𝑿) is defined as
𝐹𝐹 𝑥𝑥 = 𝑃𝑃 𝑋𝑋 ≤ 𝑥𝑥

CDF is the cumulative sum of the pmf up to 𝑥𝑥.

Properties:
1. 0 ≤ 𝐹𝐹 𝑥𝑥 ≤ 1
2. 𝐹𝐹 𝑥𝑥 is a non-decreasing function. i.e. 𝐹𝐹 𝑥𝑥2 ≥ 𝐹𝐹 𝑥𝑥1
if 𝑥𝑥2 > 𝑥𝑥1

SEHH2311 Foundations of Data Science Page 10

 Example 4
Suppose 𝑋𝑋 is the number showing up when rolling a fair
die.
1
pmf: 𝑓𝑓 𝑥𝑥 = , for 𝑥𝑥 = 1, 2, … , 6
6

𝑥𝑥
CDF: 𝐹𝐹 𝑥𝑥 = 𝑃𝑃 𝑋𝑋 ≤ 𝑥𝑥 = , for 𝑥𝑥 = 1, 2, … , 6
6
1
𝐹𝐹 1 = 𝑃𝑃 𝑋𝑋 ≤ 1 = 𝑓𝑓 1 =
6
2
𝐹𝐹 2 = 𝑃𝑃 𝑋𝑋 ≤ 2 = 𝑓𝑓 1 + 𝑓𝑓 2 =
6
…
6
𝐹𝐹 6 = 𝑃𝑃 𝑋𝑋 ≤ 6 = 𝑓𝑓 1 + 𝑓𝑓 2 + ⋯ + 𝑓𝑓 6 =
6

SEHH2311 Foundations of Data Science Page 11

 Example 4
Visualization of pmf and CDF of X

SEHH2311 Foundations of Data Science Page 12

 Example 5
Suppose the pmf of the random variable 𝑋𝑋 is defined as follows.
𝑥𝑥 0 2 5 9 10
𝑓𝑓(𝑥𝑥) 0.2 0.2 0.0 0.1 0.5

The corresponding CDF will be

𝑥𝑥 0 2 5 9 10
𝐹𝐹(𝑥𝑥) 0.2 0.4 0.4 0.5 1.0

SEHH2311 Foundations of Data Science Page 13

 Example 6
Suppose 3 cards are drawn from an ordinary deck of cards with replacement. Let
𝑋𝑋 be the number of Spade ( ) out of the 3 cards drawn. Complete the tables
below and sketch the pmf and CDF of 𝑋𝑋.

Solution: pmf CDF

𝑥𝑥 𝑥𝑥
𝑓𝑓(𝑥𝑥) 𝐹𝐹(𝑥𝑥)

SEHH2311 Foundations of Data Science Page 14

 Example 6
Suppose 3 cards are drawn from an ordinary deck of cards with replacement. Let
𝑋𝑋 be the number of Spade ( ) out of the 3 cards drawn. Complete the tables
below and sketch the pmf and CDF of 𝑋𝑋.

Solution: pmf CDF

𝑥𝑥 0 1 2 3 𝑥𝑥 0 1 2 3
𝑓𝑓(𝑥𝑥) 0.422 0.422 0.141 0.015 𝐹𝐹(𝑥𝑥) 0.422 0.844 0.985 1

SEHH2311 Foundations of Data Science Page 15

 Expectation of Discrete Random
Variables
Expectation of a random variable 𝑋𝑋 is its value on average if you
observe the random variable for a large number of times.
Expectation of 𝑋𝑋 is denoted by 𝐸𝐸(𝑋𝑋) or 𝜇𝜇 (the population mean of
𝑋𝑋).
𝜇𝜇 = 𝐸𝐸 𝑋𝑋 = � 𝑥𝑥 ⋅ 𝑓𝑓(𝑥𝑥)
𝑥𝑥∈𝑆𝑆

If the random variable 𝑋𝑋 only has non-zero probability for 𝑛𝑛 distinct

values 𝑥𝑥1 , 𝑥𝑥2 , …, 𝑥𝑥𝑛𝑛 , the expected value can be written as
𝑛𝑛
n can be infinity
𝜇𝜇 = 𝐸𝐸 𝑋𝑋 = � 𝑥𝑥𝑖𝑖 ⋅ 𝑓𝑓 𝑥𝑥𝑖𝑖 for some random
variables

𝑖𝑖=1

SEHH2311 Foundations of Data Science Page 16

 Example 7
Suppose 𝑋𝑋 is the number showing up when rolling a fair die. Find
the expected value of 𝑋𝑋.
Solution: 𝑥𝑥𝑖𝑖 1 2 3 4 5 6

𝑓𝑓(𝑥𝑥𝑖𝑖 ) 1/6 1/6 1/6 1/6 1/6 1/6

𝐸𝐸 𝑋𝑋 = � 𝑥𝑥𝑖𝑖 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 )
𝑖𝑖=1
1 1 1 1 1 1
=1 +2 +3 +4 +5 +6
6 6 6 6 6 6
= 3.5
That means the average of the numbers showing up when rolling a
fair die for a large number times is equal to 3.5.

SEHH2311 Foundations of Data Science Page 17

 Example 8
Consider the random variable in Example 2, we have two $2 coins,
three $5 coins and five $10 coins in a box. Let 𝑋𝑋 be the "value" of a
coin drawn randomly from the box.
Find the expected value of the coin drawn.
Solution:

SEHH2311 Foundations of Data Science Page 18

 Example 8
Consider the random variable in Example 2, we have two $2 coins,
three $5 coins and five $10 coins in a box. Let 𝑋𝑋 be the "value" of a
coin drawn randomly from the box.
Find the expected value of the coin drawn.
Solution:
𝑥𝑥𝑖𝑖 2 5 10
𝑓𝑓 𝑥𝑥𝑖𝑖 0.2 0.3 0.5

𝐸𝐸 𝑋𝑋 = � 𝑥𝑥𝑖𝑖 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 )
𝑖𝑖=1
= 2 0.2 + 5 0.3 + 10 0.5
= $6.9

SEHH2311 Foundations of Data Science Page 19

 Variance & Standard Deviation of
Discrete Random Variable
Variance measures the "spread" of a random variable when you
observe it for a large number of time. It is defined as
𝑛𝑛

𝜎𝜎 2 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = � 𝑥𝑥𝑖𝑖 − 𝜇𝜇 2
⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 )
𝑖𝑖=1
or equivalently
𝑛𝑛

𝜎𝜎 2 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = � 𝑥𝑥𝑖𝑖2 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 ) − 𝜇𝜇2

𝑖𝑖=1

Since variance does not have the same unit as the random variable
𝑋𝑋, we define standard deviation as the square root of the variance.

𝜎𝜎 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = 𝜎𝜎 2

SEHH2311 Foundations of Data Science Page 20

 Variance & Standard Deviation of
Discrete Random Variable
Interpretation of variance:
• 𝑥𝑥𝑖𝑖 − 𝜇𝜇 2 is the squared distance of 𝑥𝑥𝑖𝑖 from the mean

• If you define 𝑌𝑌 = 𝑋𝑋 − 𝜇𝜇 2 , 𝑉𝑉𝑉𝑉𝑉𝑉(𝑋𝑋) is equal to the

expectation of 𝑌𝑌. i.e. 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = 𝐸𝐸 𝑌𝑌 = 𝐸𝐸[ 𝑋𝑋 − 𝜇𝜇 2 ]

• If most of the 𝑥𝑥𝑖𝑖 's are far from the mean 𝜇𝜇, we will have
a large variance.

SEHH2311 Foundations of Data Science Page 21

 Example 9
Suppose 𝑋𝑋 is the number showing up when rolling a fair die. Find
the variance and standard deviation of 𝑋𝑋.
Solution:
𝑛𝑛
1 1 1
𝜇𝜇 = � 𝑥𝑥𝑖𝑖 ⋅ 𝑓𝑓 𝑥𝑥𝑖𝑖 =1 +2 + ⋯+ 6 = 3.5
6 6 6
𝑖𝑖=1
𝑛𝑛

𝜎𝜎 2 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = � 𝑥𝑥𝑖𝑖2 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 ) − 𝜇𝜇2

𝑖𝑖=1
1 1 1
= 12 +2 2 + ⋯6 2 − 3.5 ≈ 2.917
6 6 6
𝜎𝜎 = 𝜎𝜎 2 ≈ 1.078

SEHH2311 Foundations of Data Science Page 22

 Example 10
Consider the random variable in Example 2, we have two $2 coins,
three $5 coins and five $10 coins in a box. Let 𝑋𝑋 be the "value" of a
coin drawn randomly from the box.
Find the variance and standard deviation of 𝑋𝑋.
Solution:

SEHH2311 Foundations of Data Science Page 23

 Example 10
Consider the random variable in Example 2, we have two $2 coins,
three $5 coins and five $10 coins in a box. Let 𝑋𝑋 be the "value" of a
coin drawn randomly from the box.
Find the variance and standard deviation of 𝑋𝑋.
Solution:
3

𝜇𝜇 = � 𝑥𝑥𝑖𝑖 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 ) = 2 0.2 + 5 0.3 + 10 0.5 = $6.9

𝑖𝑖=1
3

𝜎𝜎 2 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = � 𝑥𝑥𝑖𝑖2 ⋅ 𝑓𝑓(𝑥𝑥𝑖𝑖 ) − 𝜇𝜇2 = 22 0.2 + 52 0.3 + 102 0.5 − 6.92 = 10.69
𝑖𝑖=1
𝜎𝜎 = 𝜎𝜎 2 ≈ $3.270

SEHH2311 Foundations of Data Science Page 24

 Continuous Random Variable
• Examples of continuous random variable
– Height of a randomly selected student from a school
– Speed of a randomly selected car on a highway
• The behavior of a continuous random variable can be fully
described by its probability density function (pdf) 𝑓𝑓(𝑥𝑥)
• Sample space (or domain) of a continuous random variable
typically consists of intervals of values instead of a set of distinct
values
• Similar to discrete random variable, we also have cumulative
distribution function (CDF) F(x) for continuous variable.

SEHH2311 Foundations of Data Science Page 25

 Continuous Random Variable
Properties of pdf
Properties of a pdf 𝑓𝑓(𝑥𝑥) for the random variable 𝑋𝑋
1. 𝑓𝑓 𝑥𝑥 ≥ 0 for all 𝑥𝑥.
∞
2. ∫−∞ 𝑓𝑓(𝑥𝑥) =1.
𝑥𝑥2
3. 𝑃𝑃 𝑥𝑥1 < 𝑋𝑋 < 𝑥𝑥2 = ∫𝑥𝑥 𝑓𝑓(𝑥𝑥) 𝑑𝑑𝑑𝑑 for any 𝑥𝑥1 less than
1
𝑥𝑥2

SEHH2311 Foundations of Data Science Page 26

 Continuous Random Variable
Cumulative Distribution Function (CDF)
Cumulative Distribution Function 𝑭𝑭(𝑿𝑿) is defined as
𝑥𝑥
𝐹𝐹 𝑥𝑥 = 𝑃𝑃 𝑋𝑋 ≤ 𝑥𝑥 = � 𝑓𝑓 𝑢𝑢 𝑑𝑑𝑑𝑑
−∞

CDF is the cumulative probability up to 𝑥𝑥.

Properties:
1. 0 ≤ 𝐹𝐹 𝑥𝑥 ≤ 1
2. 𝐹𝐹 𝑥𝑥 is a non-decreasing function. i.e. 𝐹𝐹 𝑥𝑥2 ≥ 𝐹𝐹 𝑥𝑥1
if 𝑥𝑥2 > 𝑥𝑥1
SEHH2311 Foundations of Data Science Page 27
 Uniform Distribution
A random variable 𝑋𝑋 with a uniform distribution U(a,b) has a
constant pdf in a given interval [a, b]. The pdf 𝑓𝑓(𝑥𝑥) has the
following form
1
𝑓𝑓𝑓𝑓𝑓𝑓 𝑎𝑎 ≤ 𝑥𝑥 ≤ 𝑏𝑏
𝑓𝑓 𝑥𝑥 = �𝑏𝑏 − 𝑎𝑎
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < 𝑎𝑎 𝑜𝑜𝑜𝑜 𝑥𝑥 > 𝑏𝑏
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < 𝑎𝑎
𝑥𝑥 − 𝑎𝑎
𝐹𝐹 𝑥𝑥 = 𝑓𝑓𝑓𝑓𝑓𝑓 𝑎𝑎 ≤ 𝑥𝑥 ≤ 𝑏𝑏
𝑏𝑏 − 𝑎𝑎
1 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 > 𝑏𝑏

SEHH2311 Foundations of Data Science Page 28

 Uniform Distribution
Suppose 𝑋𝑋 has the distribution 𝑈𝑈 1,4 , i.e. a=1 and b=4.
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < 1
1
𝑓𝑓𝑓𝑓𝑓𝑓 1 ≤ 𝑥𝑥 ≤ 4 𝑥𝑥 − 1
𝑓𝑓 𝑥𝑥 = �3 𝐹𝐹 𝑥𝑥 = 𝑓𝑓𝑓𝑓𝑓𝑓 1 ≤ 𝑥𝑥 ≤ 4
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < 1 𝑜𝑜𝑜𝑜 𝑥𝑥 > 4 3
1 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 > 4

SEHH2311 Foundations of Data Science Page 29

 Uniform Distribution
Suppose 𝑋𝑋 has the distribution 𝑈𝑈(1,4). Find the following probabilities.
1. P( 1.5 < 𝑋𝑋 < 2)
2. P( -1 < 𝑋𝑋 < 3)
3. P( X = 3)
Solution:
2 1 1
1. 𝑃𝑃 1.5 < 𝑋𝑋 < 2 = ∫1.5 3
𝑑𝑑𝑑𝑑 = 2 − 1.5
3
≈ 0.167
3 1 3 1
2. 𝑃𝑃 −1 < 𝑋𝑋 < 3 = ∫−1 𝑓𝑓(𝑢𝑢)𝑑𝑑𝑑𝑑 = ∫−1 0 𝑑𝑑𝑑𝑑 + ∫1 3
𝑑𝑑𝑑𝑑
1
=0+ 3−1 ≈ 0.667
3
3 1 1
3. 𝑃𝑃 𝑋𝑋 = 3 = ∫3 3
𝑑𝑑𝑑𝑑 = 3 − 3
3
=0

SEHH2311 Foundations of Data Science Page 30

 Example 11
Suppose a random variable 𝑋𝑋 has a pdf as shown below. Write
down the pdf and CDF of 𝑋𝑋.
Solution:

SEHH2311 Foundations of Data Science Page 31

 Example 11
Suppose a random variable 𝑋𝑋 has a pdf as shown below. Write
down the pdf and CDF of 𝑋𝑋.
Solution:

0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < −1
1 + 𝑥𝑥 𝑓𝑓𝑓𝑓𝑓𝑓 − 1 ≤ 𝑥𝑥 ≤ 0
𝑓𝑓 𝑥𝑥 =
1 − 𝑥𝑥 𝑓𝑓𝑓𝑓𝑓𝑓 0 < 𝑥𝑥 ≤ 1
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 > 1
0 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 < −1
𝑥𝑥 + 𝑥𝑥 2 /2 + 1/2 𝑓𝑓𝑓𝑓𝑓𝑓 − 1 ≤ 𝑥𝑥 ≤ 0
𝐹𝐹 𝑥𝑥 =
𝑥𝑥 − 𝑥𝑥 2 /2 + 1/2 𝑓𝑓𝑓𝑓𝑓𝑓 0 < 𝑥𝑥 ≤ 1
1 𝑓𝑓𝑓𝑓𝑓𝑓 𝑥𝑥 > 1

SEHH2311 Foundations of Data Science Page 32

 Example 12
For the random variable described in Example 11, find the following
probabilities.
Solution:

𝑃𝑃 𝑋𝑋 ≤ −0.5 = 𝐹𝐹 =

𝑃𝑃 𝑋𝑋 ≤ 0.2 = 𝐹𝐹 =

𝑃𝑃 −0.5 < 𝑋𝑋 ≤ 0.2 =

𝑃𝑃 0 < 𝑋𝑋 < 0.8 =

SEHH2311 Foundations of Data Science Page 33

 Example 12
For the random variable described in Example 11, find the following
probabilities.
Solution:

−0.5 2 1
𝑃𝑃 𝑋𝑋 ≤ −0.5 = 𝐹𝐹 −0.5 = −0.5 + + = 0.125
2 2

0.2 2 1
𝑃𝑃 𝑋𝑋 ≤ 0.2 = 𝐹𝐹 0.2 = 0.2 − + = 0.68
2 2

𝑃𝑃 −0.5 < 𝑋𝑋 ≤ 0.2 = 𝐹𝐹 0.2 − 𝐹𝐹 −0.5 = 0.68 − 0.125 = 0.555

𝑃𝑃 0 < 𝑋𝑋 < 0.8 = 𝐹𝐹 0.8 − 𝐹𝐹 0 = 0.98 − 0.5 = 0.48

SEHH2311 Foundations of Data Science Page 34

 Expectation of Continuous Random
Variable
Similar to discrete random variable the expectation of continuous
random variable is the average value of a random variable 𝑋𝑋 when
you observe it for large number of times. Expectation of continuous
random variable is defined as follows
∞
𝜇𝜇 = 𝐸𝐸 𝑋𝑋 = � 𝑥𝑥 ⋅ 𝑓𝑓 𝑥𝑥 𝑑𝑑𝑑𝑑
−∞

SEHH2311 Foundations of Data Science Page 35

 Variance and Standard Deviation of
Continuous Random Variable
Similar to discrete random variable, we can use the expectation of
squared distance from the mean to define variance of continuous
random variable.
∞
𝜎𝜎 2 = 𝑉𝑉𝑉𝑉𝑉𝑉(𝑋𝑋) = 𝐸𝐸[ 𝑋𝑋 − 𝜇𝜇 2 ] = � 𝑥𝑥 − 𝜇𝜇 2 ⋅ 𝑓𝑓 𝑥𝑥 𝑑𝑑𝑑𝑑
−∞
or equivalently, 𝜎𝜎 2 = 𝐸𝐸 𝑋𝑋 2 − 𝜇𝜇2

Analogously, we can define standard deviation as below

𝜎𝜎 = 𝑉𝑉𝑉𝑉𝑉𝑉 𝑥𝑥 = 𝜎𝜎 2

SEHH2311 Foundations of Data Science Page 36

 Example 13
Suppose 𝑋𝑋 has a uniform distribution 𝑈𝑈(2,4). Find the expected value and
variance of 𝑋𝑋.
Solution:
4
4 1
𝐸𝐸 𝑋𝑋 = � 𝑥𝑥 ⋅ 𝑓𝑓 𝑥𝑥 𝑑𝑑𝑑𝑑 = � 𝑥𝑥 𝑑𝑑𝑑𝑑
2 2 4 − 2
4
𝑥𝑥 2
42 22
= � = − =3
4 4 4
2
4
𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 = 𝐸𝐸 𝑋𝑋 2 − 𝜇𝜇2 = � 𝑥𝑥 2 ⋅ 𝑓𝑓 𝑥𝑥 𝑑𝑑𝑑𝑑 − 𝜇𝜇2
2
4
4
1 𝑥𝑥 3
=� 𝑥𝑥 2 ⋅ 𝑑𝑑𝑑𝑑 − 𝜇𝜇2 = � − 32 ≈ 0.333
2 4−2 6
2

SEHH2311 Foundations of Data Science Page 37

 Functions of Random Variables
In some situations, we need to work with functions of several
random variables. For example, 𝑋𝑋, 𝑌𝑌 and 𝑍𝑍 are the Chinese, English
and Mathematics scores of a randomly selected student from a
school. It will be interesting to know the expectation and variance
of the average score of the three subjects.

𝑋𝑋 + 𝑌𝑌 + 𝑍𝑍
𝐸𝐸 =?
3
𝑋𝑋 + 𝑌𝑌 + 𝑍𝑍
𝑉𝑉𝑉𝑉𝑉𝑉 =?
3

SEHH2311 Foundations of Data Science Page 38

 Useful Rules About Expectations
Suppose 𝑋𝑋 and 𝑌𝑌 are two random variables. 𝑎𝑎 and 𝑏𝑏 are
two scalars.
𝐸𝐸 𝑎𝑎𝑎𝑎 + 𝑏𝑏𝑏𝑏 = 𝑎𝑎𝑎𝑎 𝑋𝑋 + 𝑏𝑏𝑏𝑏(𝑌𝑌)

If 𝑋𝑋 and 𝑌𝑌 are independent, then

𝐸𝐸 𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑎𝑎𝑎𝑎 𝑋𝑋 𝐸𝐸(𝑌𝑌)

𝑋𝑋 𝐸𝐸 𝑋𝑋
However, 𝐸𝐸 ≠ in general!!!
𝑌𝑌 𝐸𝐸 𝑌𝑌

SEHH2311 Foundations of Data Science Page 39

 Useful Rules About Variances
Suppose 𝑋𝑋 and 𝑌𝑌 are two independent random variables.
𝑎𝑎 and 𝑏𝑏 are two scalars.

𝑉𝑉𝑉𝑉𝑉𝑉 𝑎𝑎𝑎𝑎 + 𝑏𝑏𝑏𝑏 = 𝑎𝑎2 𝑉𝑉𝑉𝑉𝑉𝑉 𝑋𝑋 + 𝑏𝑏 2 𝑉𝑉𝑉𝑉𝑉𝑉(𝑌𝑌)

Important: The above relationship does not hold if 𝑋𝑋 and

𝑌𝑌 are dependent.

SEHH2311 Foundations of Data Science Page 40