# 程序代写代做代考 data science Introduction to information system

Introduction to information system

Popular Distributions (1/2)

Bowei Chen

School of Computer Science

University of Lincoln

CMP3036M/CMP9063M Data Science

• Univariate Distributions

– Discrete Distributions

• Uniform

• Bernoulli

• Binomial

• Poisson

– Continuous Distributions

• Uniform

• Normal/Gaussian

• Exponential

• Multivariate Distributions

– Multivariate Normal Distribution

Objectives

Today’s Objectives

Popular Distributions (1/2)

Warm-up Questions:

1) What is the probability that the

dice is 2?

2) What is the probability that the

dice is less than 3?

3) If you will be given the money as

the number the dice is shown

(e.g., £1 if dice is 1) and you can

play infinite times of this game,

how much money you can expect

to earn per game?

1

6

2

6

£3.5

𝑋(𝜔1) = 1𝜔1 =

𝜔2 =

𝜔3 =

𝜔4 =

𝜔5 =

𝜔6 =

Ω

𝑋(𝜔2) = 2

𝑋(𝜔3) = 3

𝑋(𝜔4) = 4

𝑋(𝜔5) = 5

𝑋(𝜔6) = 6

ℙ 𝜔1 = ℙ 𝑋 𝜔1 = 1 =
1

6

ℙ 𝜔2 = ℙ 𝑋 𝜔2 = 2 =
1

6

ℙ 𝜔3 = ℙ 𝑋 𝜔3 = 3 =
1

6

ℙ 𝜔4 = ℙ 𝑋 𝜔4 = 4 =
1

6

ℙ 𝜔5 = ℙ 𝑋 𝜔5 = 5 =
1

6

ℙ 𝜔6 = ℙ 𝑋 𝜔6 = 6 =
1

6

Solutions:

1) What is the probability that the dice is 2?

2) What is the probability that the dice is less than 3?

3) If you will be given the money as the number the dice is (e.g., £1 if dice is

1) and you can play infinite times of this game, how much money you can

expect to earn for each game?

ℙ 𝑋(𝜔1) = 2 =
1

6

ℙ 𝑋(𝜔1) = 1 + ℙ 𝑋(𝜔2) = 2 =
2

6

£1 × ℙ 𝑋(𝜔1) = 1 + £2 × ℙ 𝑋(𝜔2) = 2 + ⋯+ £6 × ℙ 𝑋 𝜔6 = 6

= £1 +⋯+ £6 ×
1

6
= £3.5

Discrete Uniform Distribution

• Notation

𝑋~Uniform( 1,⋯ , 𝑁 )

• PMF/PDF

𝑓 𝑥 = ℙ 𝑋 = 𝑥 =
1

𝑁
, if 𝑥 = 1,⋯ ,𝑁,

0, otherwise.

• Expectation and variance

𝔼(𝑋) =
𝑁 + 1

2
,

𝕍(𝑋) =
𝑁2 − 1

12
.

Questions:

There is a flipping (fair) coin game. If head

is shown, you can earn £1; if tail is shown,

you earn nothing. How much do you think

should you be charged to enter the game?

What if head happens with probability 0.4?

£1 × 0.5 + £0 × 0.5 = £0.5

£1 × 0.4 + £0 × 0.6 = £0.4

Bernoulli Distribution

• Notation

𝑋~Bernoulli 𝑝

• PMF/PDF

𝑓 𝑥; 𝑝 = ℙ 𝑋 = 𝑥 =
𝑝𝑥(1 − 𝑝)1−𝑥, if 𝑥 = 0,1,

0, otherwise,

and 0 ≤ 𝑝 ≤ 1.

• Expectation and variance
𝔼(𝑋) = 𝑝,
𝕍(𝑋) = 𝑝(1 − 𝑝).

Questions:

An unfair coin is flipped 10 times. For each

flipping, head happens with probability 0.4.

What is the probability that exactly 6 heads

will occur?

ℙ 𝑋 = 6 =
6
10

× 0.46 × 0.64 = 0.1114767.

Choose 6 outcomes from

10 flipping outcomes

Binomial Distribution

• Notation

𝑋~Binomial 𝑛, 𝑝 or Bin 𝑛, 𝑝

• PMF/PDF

𝑓(𝑥; 𝑛, 𝑝) = ℙ 𝑋 = 𝑥 =

𝑛

𝑥
𝑝𝑥(1 − 𝑝)𝑛−𝑥, if 𝑥 = 0,⋯ , 𝑛,

0, otherwise,

and 0 ≤ 𝑝 ≤ 1.

• Expectation and variance
𝔼(𝑋) = 𝑛𝑝,
𝕍(𝑋) = 𝑛𝑝(1 − 𝑝).

Questions:

You usually receive 5 text messages per hour.

How likely that you will receive:

1) Exactly 2 messages in the coming hour?

2) 10 or fewer messages in 2 hours?

Poisson Distribution

The Poisson distribution is a discrete probability distribution for the counts of

events that occur randomly in a given interval of time (or space).

If we let 𝑋 be the number of events in a given interval, Then, if the mean
number of events per interval is 𝜆, the probability of observing 𝑥 events in a
given interval is given by

ℙ 𝑋 = 𝑥 =
𝑒−𝜆𝜆𝑥

𝑥!
, 𝑥 = 0,1,2,⋯ .

Solutions:

You usually receive 5 text messages per hour. How likely that you will receive:

1) Exactly 2 messages in the coming hour?

ℙ 𝑋 = 2 =
𝑒−552

2!
= 0.084

2) 10 or fewer messages in 2 hours?

ℙ 𝑋 ≤ 10 =

𝑥=0

10
𝑒−1010𝑥

𝑥!
= 0.5831

Poisson Distribution

• Notation

𝑋~Poisson(𝜆)

• PMF/PDF

𝑓 𝑥; 𝜆 = ℙ 𝑋 = 𝑥 =
𝑒−𝜆𝜆𝑥

𝑥!
, 𝑥 = 0,1,2,⋯ .

• Expectation and variance
𝔼(𝑋) = 𝜆,
𝕍(𝑋) = 𝜆.

Poisson Approximation of Binomial Probabilities

𝑋 = 1 𝑋 = 0

text messages per hour.

That is 𝔼 𝑋 = 𝑛𝑝 = 5 = 𝜆.
We can consider 𝑛 is very
large and 𝑝 is very small.

Recall the PMF of the Binomial distribution:

𝑓 𝑥; 𝑛, 𝑝 =
𝑛

𝑥
𝑝𝑥 1 − 𝑝 𝑛−𝑥 =

𝑛 𝑛 − 1 ⋯(𝑛 − 𝑥 + 1)

𝑥!

𝜆

𝑛

𝑥

1 −
𝜆

𝑛

𝑛−𝑥

The larger the 𝑛 and the smaller the 𝑝, the better is the approximation.

Poisson Approximation of Binomial Probabilities

𝜆𝑥

𝑥!
1 −

𝜆

𝑛

𝑛

=
𝜆𝑥

𝑥!
𝑒−𝜆 = 𝑓(𝑥; 𝜆)

References

• G.Casella, R.Berger (2002) Statistical Inference. Chapter 3

• K.Murphy (2012) Machine Learning: A Probabilistic Perspective. Chapter 2

Thank You!

bchen@Lincoln.ac.uk

mailto:bchen@Lincoln.ac.uk

Posted in Uncategorized