1 of 70

Chapter 4 Random Variables

Ver.102525

SPHsu_Probbability

1

2 of 70

SPHsu_Probbability

2

What is a random variable ?

A formal definition :

3 of 70

SPHsu_Probbability

3

Example

Therefore,

4 of 70

SPHsu_Probbability

4

Discrete Random Variables

5 of 70

SPHsu_Probbability

5

An example on the random variable value v.s. probability

6 of 70

SPHsu_Probbability

6

Example

7 of 70

SPHsu_Probbability

7

Cumulative distribution function (cdf)

8 of 70

SPHsu_Probbability

8

Cumulative distribution function (cdf)

for a discrete random variable.

9 of 70

SPHsu_Probbability

9

Expected value

Example

10 of 70

SPHsu_Probbability

10

Example

11 of 70

SPHsu_Probbability

11

Suppose the random variable X is nonnegative and integer-valued, then

This can be seen by observing that

12 of 70

SPHsu_Probbability

12

13 of 70

SPHsu_Probbability

13

Expectation of a function of a random variable

Attention !!

14 of 70

SPHsu_Probbability

14

15 of 70

SPHsu_Probbability

15

Example

16 of 70

SPHsu_Probbability

16

17 of 70

SPHsu_Probbability

17

18 of 70

SPHsu_Probbability

18

Variance

Alternatively,

19 of 70

SPHsu_Probbability

19

Example

20 of 70

SPHsu_Probbability

20

pf.

Proposition

21 of 70

Average household size

In 2011 the average household in Hong Kong had 2.9 people.

Take a random person. What is the average number of people in his/her household?

B: 2.9

A: < 2.9

C: > 2.9

22 of 70

Average household size

average�household size

3

3

average size of random�person’s household

3

4⅓

23 of 70

A general solution

24 of 70

This little mobius strip of a phenomenon is called the “generalized friendship paradox,” and at first glance it makes no sense. Everyone’s friends can’t be richer and more popular — that would just escalate until everyone’s a socialite billionaire.

The whole thing turns on averages, though. Most people have small numbers of friends and, apparently, moderate levels of wealth and happiness. A few people have buckets of friends and money and are (as a result?) wildly happy. When you take the two groups together, the really obnoxiously lucky people skew the numbers for the rest of us. Here’s how MIT’s Technology Review explains the math:

The paradox arises because numbers of friends people have are distributed in a way that follows a power law rather than an ordinary linear relationship. So most people have a few friends while a small number of people have lots of friends.

It’s this second small group that causes the paradox. People with lots of friends are more likely to number among your friends in the first place. And when they do, they significantly raise the average number of friends that your friends have. That’s the reason that, on average, your friends have more friends than you do.

And this rule doesn’t just apply to friendship — other studies have shown that your Twitter followers have more followers than you, and your sexual partners have more partners than you’ve had. This latest study, by Young-Ho Eom at the University of Toulouse and Hang-Hyun Jo at Aalto University in Finland, centered on citations and coauthors in scientific journals. Essentially, the “generalized friendship paradox” applies to all interpersonal networks, regardless of whether they’re set in real life or online.

So while it’s tempting to blame social media for what the New York Times last month called “the agony of Instagram” — that peculiar mix of jealousy and insecurity that accompanies any glimpse into other people’s glamorously Hudson-ed lives — the evidence suggests that Instagram actually has little to do with it. Whenever we interact with other people, we glimpse lives far more glamorous than our own.

That’s not exactly a comforting thought, but it should assuage your FOMO next time you scroll through your Facebook feed.

25 of 70

Bob

Mark

Zoe

Eve

Sam

Jessica

X = number of friends

Y = number of friends

of a friend

It can be shown that E[Y] ≥ E[X] in any social network . To see this, note that for a connected simple k-vertex graph, we have

Alice

where di is the valency of vertex i.

26 of 70

SPHsu_Probbability

26

Which way is more probable to find a firstborn child ? A random selection from the people in the street or a random call to any family to find ?

A random selection from the people in the street:

A random call to any family:

By Cauchy’s inequality

27 of 70

SPHsu_Probbability

27

Bernoulli distribution

Binomial distribution

28 of 70

SPHsu_Probbability

28

Example

29 of 70

SPHsu_Probbability

29

30 of 70

SPHsu_Probbability

30

31 of 70

SPHsu_Probbability

31

Properties

32 of 70

SPHsu_Probbability

32

As a result,

33 of 70

SPHsu_Probbability

33

34 of 70

SPHsu_Probbability

34

35 of 70

SPHsu_Probbability

35

Poisson distribution

It can be derived from the binomial distribution, under some extreme conditions…

36 of 70

SPHsu_Probbability

36

size

We have

37 of 70

SPHsu_Probbability

37

Note that

As a result,

38 of 70

SPHsu_Probbability

38

In the birthday problem,

39 of 70

SPHsu_Probbability

39

40 of 70

SPHsu_Probbability

40

41 of 70

SPHsu_Probbability

41

A closer look at Poisson distribution:

42 of 70

SPHsu_Probbability

42

43 of 70

SPHsu_Probbability

43

44 of 70

SPHsu_Probbability

44

As a result,

and

45 of 70

SPHsu_Probbability

45

Occurrence rate and waiting time

46 of 70

SPHsu_Probbability

46

Other property

47 of 70

SPHsu_Probbability

47

Geometric distribution

48 of 70

SPHsu_Probbability

48

As a result,

49 of 70

SPHsu_Probbability

49

As a result,

50 of 70

SPHsu_Probbability

50

Negative binomial

success.

SHOW THAT

It is well known that

?

We have more!

It is well known that

We have more!

HINTs

51 of 70

SPHsu_Probbability

51

Back to the problem

?

pf.

Since negative binomial is involved, the distribution is called the Negative binomial distribution.

Suppose

52 of 70

SPHsu_Probbability

52

Using negative binomial to prove the Chu-Vandermonde identity:

53 of 70

SPHsu_Probbability

53

Example

Compared to the solution by Fermat in Chapter 3, we have an interesting identity:

We show this in the following.

54 of 70

SPHsu_Probbability

54

We need the identity below.

To see this, we check on both sides of the equality the coefficient of , p>n. In the left hand side we have

which is the same as that in the right hand side. (Here we

recall from Ch.1 that)

Letting a=p and b=1-p yields the result in the previous page.

55 of 70

SPHsu_Probbability

55

56 of 70

SPHsu_Probbability

56

The general playoff problem in Ch.3.

Suppose team A and B meet for a (2k+1)-game playoff for some positive integer k. They play until one team wins k+1 games. If team A has a winning probability p in every game, and Pi is the probability that i games are played, show that

57 of 70

SPHsu_Probbability

57

We first prove that

thus the proof is completed.

pf.

58 of 70

SPHsu_Probbability

58

The probability that i games are played is

Finally, we have

where follows from the

previous page by letting a=p and b=1-p.

59 of 70

SPHsu_Probbability

59

Hypergeometric distribution

60 of 70

SPHsu_Probbability

60

61 of 70

SPHsu_Probbability

61

62 of 70

SPHsu_Probbability

62

ment

63 of 70

SPHsu_Probbability

63

Note that

64 of 70

SPHsu_Probbability

64

Expected values of Sums of random variables.

65 of 70

SPHsu_Probbability

65

Example

66 of 70

SPHsu_Probbability

66

More generally,

67 of 70

SPHsu_Probbability

67

Properties of cumulative distribution function:

68 of 70

SPHsu_Probbability

68

Example

We thus have

69 of 70

SPHsu_Probbability

69

Let q=1-p.

and the proof is completed.

pf:

70 of 70

SPHsu_Probbability

70

As a result,