0% found this document useful (0 votes)
160 views72 pages

Discrete Probability Distributions

The document discusses different types of probability distributions including discrete and continuous distributions. It provides examples of discrete probability distributions such as the binomial distribution and Poisson distribution. For discrete distributions, it describes their properties including that the probability function f(x) must be greater than or equal to 0, the probabilities must sum to 1, and the probability P(X=x) is equal to the value of the function f(x). It also provides examples of calculating the mean and variance for discrete uniform distributions.

Uploaded by

Paul Uy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
160 views72 pages

Discrete Probability Distributions

The document discusses different types of probability distributions including discrete and continuous distributions. It provides examples of discrete probability distributions such as the binomial distribution and Poisson distribution. For discrete distributions, it describes their properties including that the probability function f(x) must be greater than or equal to 0, the probabilities must sum to 1, and the probability P(X=x) is equal to the value of the function f(x). It also provides examples of calculating the mean and variance for discrete uniform distributions.

Uploaded by

Paul Uy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 72

Discrete

Probability Distributions
Types of Probability Distribution

1. Discrete probability distribution - one that involves a discrete


random variable.
• Example: Poisson Distribution (x is the no. of arrivals per time period)

2. Continuous probability distribution - one that involves a


continuous random variable.
• Example: Exponential Distribution (t is the inter-arrival time of customers)
Probability Distribution

• Probability Distribution is a table or a function which helps us


determine or compute the probability associated to each value of the
random variable.
Discrete Probability Distribution

Characteristics of a discrete probability distribution


1. f(x)  0 x (for all x)
2. f(x) = 1
3. P(X = x) = f(x) → refers to the value of the function when the
random variable X is equal to a specific value x.
• Ex. In the experiment of tossing a coin three times:
where X is the number of heads:
P(X = 2) = f(2) = 3/8
Discrete Probability Distribution
Example: Tossing 3 coins
Outcomes TTT, HTT, THT, TTH, HHT, HTH, THH, HHH
X= # of Heads 0 1 1 1 2 2 2 3
Properties
X= # of P (X=x)
Heads f(x) 1. f(x)  0 x (for all x)
2. f(x) = 1 ; 1/8 + 3/8 + 3/8 + 1/8 =1
0 1 1/8 3. P(X = x) = f(x) → refers to the value of the function when the random
1 3 3/8 variable X is equal to a specific value x.
2 3 3/8
P(X=0) = 1/8
3 1 1/8 P(X=1) = 3/8
P(X=2) = 3/8
P(X=3) = 1/8
Discrete Probability Distribution

• Cumulative Distribution Function


• a table or a function that determines the probability that the random variable
X takes on values that are less than or equal to a specific value x
• denoted by:

• where: L = lower limit of possible x values


Example

x f(x) F(x)
0 1/8 1/8
1 3/8 4/8
2 3/8 7/8
3 1/8 1

F(0) = P (X<0)= 1/8


F(1) = P (X<1) = f(0) + f(1) = 1/8 + 3/8 = 4/8
F(2) = P(X<2) = f(0) + f(1) + f(2) = 1/8 + 3/8 + 3/8 = 7/8
F(3) = P(X<3) = f(0) + f(1) + f(2) + f(3) = 1/8 + 3/8 + 3/8 +1/8 = 8/8=1
Discrete Probability Distributions

Discrete Uniform Distribution


Binomial Distribution
Hypergeometric Distribution
Poisson Distribution
Here is the discrete uniform random variable…

If all the discrete values of the random variable 𝑿 have an equal chance of occurring,
𝑿 is said to be discrete uniform with pmf:
𝟏
𝒑𝑿 𝒌 = , 𝒊𝒇 𝒌 = 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 , … , 𝒙𝒏−𝟐 , 𝒙𝒏−𝟏 , 𝒙𝒏
𝒏

and with:
σ𝒏
𝒊=𝟏 𝒙𝒊
𝝁 = 𝑬(𝑿) = (mean)
𝒏

𝟐 σ𝒏
𝒊=𝟏(𝒙𝒊 −𝝁)
𝟐
𝝈 = 𝑽(𝑿) = (variance)
𝒏

It is a probability distribution where all outcomes are equally likely.


Let’s do some examples.

Example 1 :
Let 𝑋 represent a random variable taking on the possible values of 0, 1, 2, 3, 4,
5 ,6, 7, 8, 9, and each of these values has equal probability. Then 𝑋 is discrete
uniformly distributed and the probability associated with each of the values is:
1
𝑃 𝑋 = 𝑥𝑖 = 𝑓 𝑥𝑖 = = 0.10
10
Let’s do a simple example. 🎯

Let 𝑿 be a number selected at random from the set of numbers {𝟓𝟏, 𝟓𝟐, … , 𝟏𝟎𝟎}. Find
the mean and the variance of 𝑿.

𝑿 is a discrete uniform random variable.


𝟏
𝒑𝑿 (𝒙) = , 𝒊𝒇 𝒙 = 𝟓𝟏, 𝟓𝟐, … , 𝟏𝟎𝟎
𝟓𝟎

𝟓𝟏 + 𝟓𝟐 + 𝟓𝟑 + ⋯ + 𝟗𝟖 + 𝟗𝟗 + 𝟏𝟎𝟎
𝝁=𝑬 𝑿 = = 𝟕𝟓. 𝟓
𝟓𝟎
𝟐 𝟐 𝟐
𝟓𝟏 − 𝟕𝟓. 𝟓 + 𝟓𝟐 − 𝟕𝟓. 𝟓 + ⋯ + 𝟏𝟎𝟎 − 𝟕𝟓. 𝟓
𝝈𝟐 = 𝑽 𝑿 = = 𝟐𝟎𝟖. 𝟐𝟓
𝟓𝟎
Let’s do another simple example. 🎯

Let 𝑿 be a number selected at random from the set of first 𝒏 positive integers. Let
𝒀 = 𝑿 + 𝒌 for any real number 𝒌. Show that 𝑬(𝒀) = 𝑬(𝑿) + 𝒌, and 𝑽(𝒀) = 𝑽(𝑿).

𝑿 and 𝒀 are discrete uniform random variables.

𝟏 + 𝟐 + 𝟑 + ⋯ + 𝒏 − 𝟏 + 𝒏 σ𝒏𝒊=𝟏 𝒊
𝝁𝑿 = 𝑬 𝑿 = =
𝒏 𝒏
(𝟏 + 𝒌) + (𝟐 + 𝒌) + (𝟑 + 𝒌) + ⋯ + (𝒏 + 𝒌) σ𝒏𝒊=𝟏 𝒊 + 𝒏𝒌
𝝁𝒀 = 𝑬 𝒀 = =
𝒏 𝒏
σ𝒏𝒊=𝟏 𝒊
𝝁𝒀 = 𝑬 𝒀 = + 𝒌 = 𝝁𝑿 + 𝒌
𝒏
Let’s do another simple example. 🎯

Let 𝑿 be a number selected at random from the set of first 𝒏 positive integers. Let
𝒀 = 𝑿 + 𝒌 for any real number 𝒌. Show that 𝑬(𝒀) = 𝑬(𝑿) + 𝒌, and 𝑽(𝒀) = 𝑽(𝑿).

𝑿 and 𝒀 are discrete uniform random variables.


𝟐 𝟐 𝟐 𝟐 𝒏 𝟐
(𝟏 − 𝝁 𝑿 ) +(𝟐 − 𝝁 𝑿 ) +(𝟑 − 𝝁 𝑿 ) + ⋯ + (𝒏 − 𝝁 𝑿 ) σ 𝒊=𝟏 (𝒊 − 𝝁 𝑿 )
𝝈𝟐𝑿 = 𝑽 𝑿 = =
𝒏 𝒏
[(𝟏 + 𝒌) − 𝝁𝒀 ]𝟐 +[(𝟐 + 𝒌) − 𝝁𝒀 ]𝟐 + ⋯ + [(𝒏 + 𝒌) − 𝝁𝒀 ]𝟐 σ𝒏𝒊=𝟏[(𝒊 + 𝒌) − 𝝁𝒀 ]𝟐
𝝈𝟐𝒀 =𝑽 𝒀 = =
𝒏 𝒏

𝟐 +[(𝟐 + 𝒌) − (𝝁 +𝒌)]𝟐 + ⋯ + [(𝒏 + 𝒌) − (𝝁 +𝒌)]𝟐 σ𝒏 𝟐


[(𝟏 + 𝒌) − (𝝁 𝑿 +𝒌)] 𝑿 𝑿 𝒊=𝟏[(𝒊 + 𝒌) − (𝝁 𝑿 +𝒌)]
𝝈𝟐𝒀 = 𝑽 𝒀 = =
𝒏 𝒏

𝒏 𝟐
σ 𝒊=𝟏 (𝒊 − 𝝁 𝑿 )
𝝈𝟐𝒀 = 𝑽 𝒀 = = 𝝈𝟐𝑿
𝒏
It just means that whenever there are shifts in the
original values, only the mean changes.
Let 𝑿 be a number selected at random from the set of first 𝒏 positive integers. Let
𝒀 = 𝑿 + 𝒌 for any real number 𝒌. Show that 𝑬(𝒀) = 𝑬(𝑿) + 𝒌, and 𝑽(𝒀) = 𝑽(𝑿).

𝑿 and 𝒀 are discrete uniform random variables.

σ𝒏𝒊=𝟏 𝒊
𝝁𝒀 = 𝑬 𝒀 = + 𝒌 = 𝝁𝑿 + 𝒌
𝒏

𝒏 𝟐
σ 𝒊=𝟏 (𝒊 − 𝝁 𝑿 )
𝝈𝟐𝒀 = 𝑽 𝒀 = = 𝝈𝟐𝑿
𝒏
And that is true for all random variables.

Let 𝑿 be a number selected at random from the set of first 𝒏 positive integers. Let
𝒀 = 𝑿 + 𝒌 for any real number 𝒌. Show that 𝑬(𝒀) = 𝑬(𝑿) + 𝒌, and 𝑽(𝒀) = 𝑽(𝑿).

𝑿 and 𝒀 are discrete uniform random variables.

σ𝒏𝒊=𝟏 𝒊
𝝁𝒀 = 𝑬 𝒀 = + 𝒌 = 𝝁𝑿 + 𝒌
𝒏

𝒏 𝟐
σ 𝒊=𝟏 (𝒊 − 𝝁 𝑿 )
𝝈𝟐𝒀 = 𝑽 𝒀 = = 𝝈𝟐𝑿
𝒏
Example

• What is the probability distribution of tossing a die?

X 1 2 3 4 5 6
P(X=x) 1/6 1/6 1/6 1/6 1/6 1/6
Example

• An employee is selected from a staff of 10 to supervise a certain


project by selecting a tag at random from a box containing 10 tags
numbered from 1 to 10. Find the probability distribution of X
representing the number on the tag that is drawn. What is the
probability that the number drawn is less than 4?

• Find the mean and variance of X


𝑋 is uniformly distributed with pmf 𝑓(𝑥)=1/10 for 𝑥=1, 2, 3, …, 10.

𝑃(𝑋<4)=𝑓(1)+𝑓(2)+𝑓(3)=3(1/10)=0.3

The mean of 𝑋 is 𝜇=1/10 (1+2+⋯+10)=5.5

The variance of 𝑋 is 𝜎2=1/10 [(1−5.5)2+(2−5.5) 2+⋯+(10−5.5) 2 ]=8.25


Binomial Distribution

Characteristics
1. The experiment consists of n repeated trials. Each trial is called a Bernoulli
trial.
2. For each trial, there are only two mutually exclusive outcomes, success and
failure. The event that corresponds to success is chosen arbitrarily. Usually, it
is defined in accordance to what is required in the problem.
Let p = probability of success
q = probability of failure
where: p + q = 1
3. The probability of success remains constant from trial to trial. The repeated
trials are independent. This means that the outcomes on any given trial does
not affect the outcomes of the succeeding trials.
Here is the binomial random variable…

Suppose there are 𝒏 independent trials in an experiment, each having probability 𝒑 for
a “success” to occur and probability 𝟏 − 𝒑 for a “failure” to occur. Let 𝑿 count the
number of “successes” in all the trials. 𝑿 is said to be binomial with pmf:
𝒏 𝒌
𝒑𝑿 𝒌 = 𝒑 (𝟏 − 𝒑)𝒏−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒌

and with:
𝝁 = 𝑬(𝑿) = 𝒏𝒑

𝝈𝟐 = 𝑽(𝑿) = 𝒏𝒑(𝟏 − 𝒑)
In the pmf, it just means that we choose the 𝒙
“successes” in all the 𝒏 trials…
Suppose there are 𝒏 independent trials in an experiment, each having probability 𝒑 for
a “success” to occur and probability 𝟏 − 𝒑 for a “failure” to occur. Let 𝑿 count the
number of “successes” in all the trials. 𝑿 is said to be binomial with pmf:
𝒏 𝒌
𝒑𝑿 𝒌 = 𝒑 (𝟏 − 𝒑)𝒏−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒌

and with:
𝝁 = 𝑬(𝑿) = 𝒏𝒑

𝝈𝟐 = 𝑽(𝑿) = 𝒏𝒑(𝟏 − 𝒑)
Then, each of the 𝒌 “successes” has a probability of
𝒑, while the rest has a probability 𝟏 − 𝒑.
Suppose there are 𝒏 independent trials in an experiment, each having probability 𝒑 for
a “success” to occur and probability 𝟏 − 𝒑 for a “failure” to occur. Let 𝑿 count the
number of “successes” in all the trials. 𝑿 is said to be binomial with pmf:
𝒏 𝒌
𝒑𝑿 𝒌 = 𝒑 (𝟏 − 𝒑)𝒏−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒌

and with:
𝝁 = 𝑬(𝑿) = 𝒏𝒑

𝝈𝟐 = 𝑽(𝑿) = 𝒏𝒑(𝟏 − 𝒑)
The mean of 𝑿 “successes” is interestingly
the 100 𝒑% of all the 𝒏 trials…
Suppose there are 𝒏 independent trials in an experiment, each having probability 𝒑 for
a “success” to occur and probability 𝟏 − 𝒑 for a “failure” to occur. Let 𝑿 count the
number of “successes” in all the trials. 𝑿 is said to be binomial with pmf:
𝒏 𝒌
𝒑𝑿 𝒌 = 𝒑 (𝟏 − 𝒑)𝒏−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒌

and with:
𝝁 = 𝑬(𝑿) = 𝒏𝒑

𝝈𝟐 = 𝑽(𝑿) = 𝒏𝒑(𝟏 − 𝒑)
Let’s do a simple example. 👷

In the last civil engineering licensure exam, DLSU was noted to have a passing rate of
59.26%. Suppose we have 20 examinees in the coming licensure exam this year. Let
𝑪 be the number of students who will pass this year. What are the probabilities that
(a) everyone will pass, and (b) at least 1 student will pass?

𝑪 is a binomial random variable. (𝑪~𝑩𝒊𝒏(𝒏 = 𝟐𝟎, 𝒑 = 𝟎. 𝟓𝟗𝟐𝟔))

𝟐𝟎
𝒑𝑪 𝒌 = (𝟎. 𝟓𝟗𝟐𝟔)𝒌 (𝟎. 𝟒𝟎𝟕𝟒)𝟐𝟎−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝟐𝟎
𝒌
𝟐𝟎
a) 𝒑𝑪 𝟐𝟎 = 𝑷 𝑪 = 𝟐𝟎 = (𝟎. 𝟓𝟗𝟐𝟔)𝟐𝟎 (𝟎. 𝟒𝟎𝟕𝟒)𝟎 = 𝟎. 𝟎𝟎𝟐𝟗%
𝟐𝟎
𝟐𝟎
b) 𝑷 𝑪 ≥ 𝟏 = 𝟏 − 𝑷 𝑪 = 𝟎 = 𝟏 − 𝟎. 𝟓𝟗𝟐𝟔 𝟎 𝟎. 𝟒𝟎𝟕𝟒 𝟐𝟎 = 𝟗𝟗. 𝟗𝟗𝟗𝟗𝟗𝟖𝟒𝟏%
𝟎
Let’s do another simple example. 💡

Two lighting systems are being proposed for an employee work area. One requires
fifty bulbs, each having a probability 0.05 of burning out within a month’s time. The
second has one hundred bulbs, each with a 0.02 burnout probability. Whichever
system is installed will be inspected once a month for the purpose of replacing
burned-out bulbs. If we compare the probabilities that each will require at least one
bulb to be replaced each month, which system is more likely to require less
maintenance?
𝑨 and 𝑩 are binomial random variables counting the number of burned-out light bulbs
in each system. (𝑨~𝑩𝒊𝒏(𝒏 = 𝟓𝟎, 𝒑 = 𝟎. 𝟎𝟓), 𝑩~𝑩𝒊𝒏(𝒏 = 𝟏𝟎𝟎, 𝒑 = 𝟎. 𝟎𝟐))

𝟓𝟎
𝒑𝑨 𝒌 = (𝟎. 𝟎𝟓)𝒌 (𝟎. 𝟗𝟓)𝟓𝟎−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝟓𝟎
𝒌
𝟏𝟎𝟎
𝒑𝑩 𝒌 = (𝟎. 𝟎𝟐)𝒌 (𝟎. 𝟗𝟖)𝟏𝟎𝟎−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝟏𝟎𝟎
𝒌
Let’s do another simple example. 💡

Two lighting systems are being proposed for an employee work area. One requires
fifty bulbs, each having a probability 0.05 of burning out within a month’s time. The
second has one hundred bulbs, each with a 0.02 burnout probability. Whichever
system is installed will be inspected once a month for the purpose of replacing
burned-out bulbs. If we compare the probabilities that each will require at least one
bulb to be replaced each month, which system is more likely to require less
maintenance?
𝑨 and 𝑩 are binomial random variables counting the number of burned-out light bulbs
in each system. (𝑨~𝑩𝒊𝒏(𝒏 = 𝟓𝟎, 𝒑 = 𝟎. 𝟎𝟓), 𝑩~𝑩𝒊𝒏(𝒏 = 𝟏𝟎𝟎, 𝒑 = 𝟎. 𝟎𝟐))

𝟓𝟎 𝟎 𝟓𝟎−𝟎
𝑷 𝑨≥𝟏 =𝟏−𝑷 𝑨=𝟎 =𝟏− 𝟎. 𝟎𝟓 𝟎. 𝟗𝟓 = 𝟗𝟐. 𝟑𝟎𝟓𝟓%
𝟎
𝟏𝟎𝟎
𝑷 𝑩≥ 𝟏 = 𝟏−𝑷 𝑩=𝟎 =𝟏− (𝟎. 𝟎𝟐)𝟎 (𝟎. 𝟗𝟖)𝟏𝟎𝟎−𝟎 = 𝟖𝟔. 𝟕𝟑𝟖𝟎%
𝟎
It seems the second system doesn’t need much
maintenance then…
Two lighting systems are being proposed for an employee work area. One requires
fifty bulbs, each having a probability 0.05 of burning out within a month’s time. The
second has one hundred bulbs, each with a 0.02 burnout probability. Whichever
system is installed will be inspected once a month for the purpose of replacing
burned-out bulbs. If we compare the probabilities that each will require at least one
bulb to be replaced each month, which system is more likely to require less
maintenance?
𝑨 and 𝑩 are binomial random variables counting the number of burned-out light bulbs
in each system. (𝑨~𝑩𝒊𝒏(𝒏 = 𝟓𝟎, 𝒑 = 𝟎. 𝟎𝟓), 𝑩~𝑩𝒊𝒏(𝒏 = 𝟏𝟎𝟎, 𝒑 = 𝟎. 𝟎𝟐))

𝟓𝟎 𝟎 𝟓𝟎−𝟎
𝑷 𝑨≥𝟏 =𝟏−𝑷 𝑨=𝟎 =𝟏− 𝟎. 𝟎𝟓 𝟎. 𝟗𝟓 = 𝟗𝟐. 𝟑𝟎𝟓𝟓%
𝟎
𝟏𝟎𝟎
𝑷 𝑩≥ 𝟏 = 𝟏−𝑷 𝑩=𝟎 =𝟏− (𝟎. 𝟎𝟐)𝟎 (𝟎. 𝟗𝟖)𝟏𝟎𝟎−𝟎 = 𝟖𝟔. 𝟕𝟑𝟖𝟎%
𝟎
Example

• The probability that a certain kind of component will survive a shock


test is 3/4.
• Find the probability that exactly 2 of the next 4 components tested
survive.
𝒏 𝒌
𝒑𝑿 𝒌 = 𝒑 (𝟏 − 𝒑)𝒏−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒌

𝒑𝑿 𝟐 =

= .2109375
Example
• The probability that a patient recovers from a rare blood disease is
0.4. If 15 people are known to have contracted this disease, what is
the probability that (a) at least 10 survive, (b) from 3 to 8 survive, and
(c) exactly 5 survive?
𝑋 is a binomial random variable that counts the number of patients that survive the
disease. 𝑋~𝐵𝑖𝑛(𝑛 = 15, 𝑝 = 0.4)
a. 𝑃 𝑋 ≥ 10 = 1 − 𝑃 𝑋 ≤ 9
= 1 − 0.9662 = 0.0338
b. 𝑃 3 ≤ 𝑋 ≤ 8
= 𝑃 𝑋 ≤ 8 − 𝑃(𝑋 ≤ 2)
= 0.9050 – 0.0271
= 0.8779
c. 𝑃 𝑋 = 5
= 𝑃 𝑋 ≤ 5 − 𝑃(𝑋 ≤ 4)
= 0.4032 – 0.2173
= 0.1859
Example

Find the mean and variance

µ = np= (15)(0.4) = 6 patients;


𝜎2 = npq= (15)(.4)(0.6) = 3.6 patients
Example

a. A large chain retailer purchases a certain kind of electronic device from a manufacturer. The
manufacturer indicates that the defective rate of the device is 3%. The inspector of the retailer
randomly picks 20 items from a shipment. What is the probability that there will be at least one
defective item among these 20? Find the mean and variance.

𝝁 = 𝑬(𝑿) = 𝒏𝒑 𝝈𝟐 = 𝑽 𝑿 = 𝒏𝒑 𝟏 − 𝒑
= 20x.03 =.60 = 20x .03 x .97=.582
Example (cont)

b. Suppose that the retailer receives 10 shipments in a month and the


inspector randomly tests 20 devices per shipment. What is the
probability that there will be 3 shipments containing at least one
defective device?

𝑌 is a binomial random variable that counts the number of


shipments with at least one defective device.
𝑌~𝐵𝑖𝑛(𝑛 = 10, 𝑝 = 0.4562)
Hypergeometric Distribution

Characteristics
1. A random sample of size n is selected from N items.
2. The N items may be subdivided into two groups, k of the items are
classified as successes. Thus, N – k of the items are considered
failures. The choice of successes is arbitrary.
3. Sampling is done without replacement.
Here is the hypergeometric random variable…

Suppose there are 𝑵 items in which 𝒌 of them are “successes”. Suppose also that we
select 𝒏 items, without replacement, from the 𝑵. Let 𝑿 count the number of
“successes” in the selected 𝒏 items. 𝑿 is said to be hypergeometric with pmf:
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒏
and with:
𝒌
𝝁 = 𝑬(𝑿) = 𝒏
𝑵

𝟐
𝒌 𝒌 𝑵−𝒏
𝝈 = 𝑽(𝑿) = 𝒏 𝟏−
𝑵 𝑵 𝑵−𝟏
In the pmf, it just means that we choose the 𝒙 from
all the 𝒏 “successes”; the rest from the “failures”.
Suppose there are 𝑵 items in which 𝒌 of them are “successes”. Suppose also that we
select 𝒏 items, without replacement, from the 𝑵. Let 𝑿 count the number of
“successes” in the selected 𝒏 items. 𝑿 is said to be hypergeometric with pmf:
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒏
and with:
𝒌
𝝁 = 𝑬(𝑿) = 𝒏
𝑵

𝟐
𝒌 𝒌 𝑵−𝒏
𝝈 = 𝑽(𝑿) = 𝒏 𝟏−
𝑵 𝑵 𝑵−𝟏
Then, the denominator is all the possible
𝒏 items from the 𝑵.
Suppose there are 𝑵 items in which 𝒌 of them are “successes”. Suppose also that we
select 𝒏 items, without replacement, from the 𝑵. Let 𝑿 count the number of
“successes” in the selected 𝒏 items. 𝑿 is said to be hypergeometric with pmf:
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒏
and with:
𝒌
𝝁 = 𝑬(𝑿) = 𝒏
𝑵

𝟐
𝒌 𝒌 𝑵−𝒏
𝝈 = 𝑽(𝑿) = 𝒏 𝟏−
𝑵 𝑵 𝑵−𝟏
Suppose there are 𝑵 items in which 𝒌 of them are “successes”. Suppose also that we
select 𝒏 items, without replacement, from the 𝑵. Let 𝑿 count the number of
“successes” in the selected 𝒏 items. 𝑿 is said to be hypergeometric with pmf:
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒏
and with:
𝒌
𝝁 = 𝑬(𝑿) = 𝒏
𝑵

𝟐
𝒌 𝒌 𝑵−𝒏
𝝈 = 𝑽(𝑿) = 𝒏 𝟏−
𝑵 𝑵 𝑵−𝟏
The only difference is in the variance because of
the correction factor.
Suppose there are 𝑵 items in which 𝒌 of them are “successes”. Suppose also that we
select 𝒏 items, without replacement, from the 𝑵. Let 𝑿 count the number of
“successes” in the selected 𝒏 items. 𝑿 is said to be hypergeometric with pmf:
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
𝒏
and with:
𝒌
𝝁 = 𝑬(𝑿) = 𝒏
𝑵

𝟐
𝒌 𝒌 𝑵−𝒏
𝝈 = 𝑽(𝑿) = 𝒏 𝟏−
𝑵 𝑵 𝑵−𝟏
Let’s do a simple example. 💉

A city has 4050 children under the age of ten, including 514 who have not been
vaccinated for measles. Sixty five of the city’s children are enrolled in the ABC Day
Care Center. Suppose the municipal health department sends a doctor and a nurse to
ABC to immunize any child who has not already been vaccinated. Find a formula for
the probability that exactly 𝒙 of the children at ABC have not been vaccinated.
𝒌 𝑵−𝒌
N=4050 𝒙 𝒏−𝒙
K=514
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
n=65 𝒏
X=x 𝟓𝟏𝟒 𝟒𝟎𝟓𝟎−𝟓𝟏𝟒
𝒙 𝟔𝟓−𝒙
𝒑𝑿 𝒙 =
𝟒𝟎𝟓𝟎
𝟔𝟓
Let’s do another simple example. 🥤

Every hour, 10 000 cans of soda are filled by a machine, among which 300 underfilled
cans are produced. Each hour, a sample of 30 cans is randomly selected, and the
number of ounces of soda per can is checked. Denote by 𝑿 the number of cans
selected that are underfilled. Find the probability that at least one underfilled can will
be among those sampled.
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
N=10000 𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
K=300 𝒏
n=30 P(X>1)= P(X<1) = 1-P(X=0)
X=1 𝟑𝟎𝟎 𝟏𝟎𝟎𝟎𝟎−𝟑𝟎𝟎
𝟎 𝟑𝟎−𝟎
𝒑𝑿 𝟎 = = .40067
𝟏𝟎𝟎𝟎𝟎
𝟑𝟎

1- = .40067 = .599
Example

• A particular part that is used as an injection device is sold in lots of


10. The producer feels that the lot is deemed acceptable if no more
than one defective is in the lot. Some lots are sampled and the
sampling plan involves random sampling and testing 3 of the parts
out of 10. If none of the 3 are defective, the lot is accepted. What is
the probability of accepting the lot?
Example

• Lots of 40 components each are called unacceptable if they contain as


many as 3 defectives or more. The procedure for sampling the lot is to
select 5 components at random and to reject the lot if a defective is
found. What is the probability that exactly 1 defective is found in the
sample if there are 3 defectives in the entire lot?
• Find the mean and variance
Example

• Find the mean and variance

5x3
µ = 40 σ2 = 40 – 5 x 5 x 3 ( 1 - 3 )
40-1 40 40
= .375

= .311
Hypergeometric Distribution

Cases where Binomial Can Approximate the Hypergeometric


Distribution
1. when N is not given (assume N is large)
2. N is given but it is very large
Example

• A manufacturer of automobile tires reports that among a shipment of


5000 sent to a local distributor, 1000 are slightly blemished. If one
purchases 10 of these tires at random from the distributor, what, is
the probability that exactly 3 are blemished?
𝒌 𝑵−𝒌
𝒙 𝒏−𝒙
𝒑𝑿 𝒙 = 𝑵
, 𝒊𝒇 𝒙 = 𝟎, 𝟏, 𝟐, … , 𝒏
N=5000
K=1000 𝒏
n=10 𝟏𝟎𝟎𝟎 𝟓𝟎𝟎𝟎−𝟏𝟎𝟎𝟎
𝟑 𝟏𝟎−𝟑
X=3 𝒑𝑿 𝟑 = 𝟓𝟎𝟎𝟎 = .2014
𝟏𝟎
Here is the Poisson random variable…

Suppose the number of outcomes occurring in one interval or region of space is


independent of that in others. Suppose also that the probability of a single outcome
occurring is the same for any two intervals or regions of equal length. Let 𝑿 count the
number of instances that will occur in a specified interval or region 𝒕, with an average
of 𝝀 outcomes per unit of time or region. 𝑿 is said to be Poisson with pmf:
𝒆−𝝀𝒕 (𝝀𝒕)𝒌
𝒑𝑿 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!

and with:
𝝁 = 𝑬(𝑿) = 𝝀𝒕

𝝈𝟐 = 𝑽(𝑿) = 𝝀𝒕

𝒆 = 𝟐. 𝟕𝟏𝟖𝟐𝟖 …
It makes sense that this is the average of 𝑿 since 𝝀
outcomes per unit done in 𝒕 periods yields 𝝀𝒕…
Suppose the number of outcomes occurring in one interval or region of space is
independent of that in others. Suppose also that the probability of a single outcome
occurring is the same for any two intervals or regions of equal length. Let 𝑿 count the
number of instances that will occur in a specified interval or region 𝒕, with an average
of 𝝀 outcomes per unit of time or region. 𝑿 is said to be Poisson with pmf:
𝒆−𝝀𝒕 (𝝀𝒕)𝒌
𝒑𝑿 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!

and with:
𝝁 = 𝑬(𝑿) = 𝝀𝒕

𝝈𝟐 = 𝑽(𝑿) = 𝝀𝒕

𝒆 = 𝟐. 𝟕𝟏𝟖𝟐𝟖 …
Let’s do a simple example. ⚡️

In a new fiber-optic communication system, transmission errors occur at the rate of


1.5 per ten seconds. What is the probability that exactly three errors will occur
during the next half-minute?

Let 𝑬 be the the Poisson random variable that counts the transmission errors in the
𝟏.𝟓
next 30 seconds. (𝑬~𝑷𝒐𝒊𝒔𝒔𝒐𝒏(𝝀𝒕 = 𝟏𝟎 ∗ 𝟑𝟎 = 𝟒. 𝟓))
𝒆−𝟒.𝟓 (𝟒. 𝟓)𝒌
𝒑𝑬 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!
𝒆−𝟒.𝟓 (𝟒. 𝟓)𝟑
𝒑𝑬 𝟑 = = 𝟏𝟔. 𝟖𝟕𝟏𝟖%
𝟑!
Let’s do a simple example. 💥

Flaws in metal sheeting produced by a high-temperature roller occur at the rate of one
per ten square feet. What is the probability that three or more flaws will appear in a
five-by-eight-foot panel?

Let 𝑭 be the the Poisson random variable that counts the flaws in a five-by-eight-foot
𝟏
panel. (𝑭~𝑷𝒐𝒊𝒔𝒔𝒐𝒏(𝝀𝒕 = 𝟏𝟎 ∗ 𝟒𝟎 = 𝟒))
𝒆−𝟒 (𝟒)𝒌
𝒑𝑭 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!
𝟐
𝒆−𝟒 𝟒 𝒌
𝑷 𝑭 ≥ 𝟑 = 𝟏 − 𝑷 𝑭 < 𝟑 = 𝟏 − 𝑷(𝑭 ≤ 𝟐) = 𝟏 − ෍ = 𝟕𝟔. 𝟏𝟖𝟗𝟕%
𝒌!
𝒌=𝟎
Example

• During a laboratory experiment the average number of radioactive


particles passing through a counter in 1 millisecond is 4. What is the
probability that 6 particles enter the counter in a given millisecond?
(𝑭~𝑷𝒐𝒊𝒔𝒔𝒐𝒏(𝝀𝒕 = 𝟒))

𝒆−𝟒 (𝟒)𝟔
𝒑𝑭 𝟔 = = . 𝟏𝟎𝟒
𝟔!
Example

• Ten is the average number of oil tankers arriving each day at a certain
port city. The facilities at the port can handle at most 15 tankers per
day. What is the probability that on a given day tankers have to be
turned away?
Since this is the limit of the binomial pmf as 𝒏 → ∞
and success probability 𝒑 → 𝟎…
Suppose the number of outcomes occurring in one interval or region of space is
independent of that in others. Suppose also that the probability of a single outcome
occurring is the same for any two intervals or regions of equal length. Let 𝑿 count the
number of instances that will occur in a specified interval or region 𝒕, with an average
of 𝝀 outcomes per unit of time or region. 𝑿 is said to be Poisson with pmf:
𝒆−𝝀𝒕 (𝝀𝒕)𝒌
𝒑𝑿 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!

and with:
𝝁 = 𝑬(𝑿) = 𝝀𝒕

𝝈𝟐 = 𝑽(𝑿) = 𝝀𝒕
We can approximate a binomial pmf with a very
large 𝒏 and a very small 𝒑 using a Poisson pmf.
Suppose 𝑿 is a binomial random variable with sample size 𝒏 and “success” probability
𝒑. If 𝒏 ≥ 𝟏𝟎𝟎 and 𝒑 < 𝟎. 𝟎𝟏, then 𝑿 is approximately Poisson with 𝝀 = 𝒏𝒑.

𝒆−𝒏𝒑 (𝒏𝒑)𝒌
𝒑𝑿 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!
Let’s do a simple example. 🧬

A chromosome mutation linked with colorblindness is known to occur, on the average,


once in every ten thousand births. Approximate the probability that exactly three of
the next twenty thousand babies born will have the mutation.

Let 𝑴 be the binomial random variable that counts the number of mutations in the
𝟏
next twenty thousand born babies. (𝑴~𝑩𝒊𝒏 𝒏 = 𝟐𝟎 𝟎𝟎𝟎, 𝒑 = 𝟏𝟎 𝟎𝟎𝟎 )
𝟏
Since 𝒏 = 𝟐𝟎 𝟎𝟎𝟎 ≥ 𝟏𝟎𝟎 and 𝒑 = < 𝟎. 𝟎𝟏, 𝑴 is approximately Poisson.
𝟏𝟎 𝟎𝟎𝟎
𝟏
(𝑴~𝑷𝒐𝒊𝒔𝒔𝒐𝒏 𝝀 = 𝒏𝒑 = 𝟐𝟎 𝟎𝟎𝟎 ∗ 𝟏𝟎 𝟎𝟎𝟎 = 𝟐 )
−𝟐 𝒌
𝒆 (𝟐)
𝒑𝑴 𝒌 = , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, …
𝒌!
𝒆−𝟐 (𝟐)𝟑
𝒑𝑴 𝟑 = = 𝟏𝟖. 𝟎𝟒𝟒𝟕%
𝟑!
Let’s do a simple example. 💊

A medical study recently documented that 905 mistakes were made among the
289 411 prescriptions written during one year at a large metropolitan teaching
hospital. Suppose a patient is admitted with a condition serious enough to warrant
200 prescriptions. Approximate the probability that at least one will contain an error.
Let 𝑿 be the binomial random variable that counts erroneous prescriptions.
𝟗𝟎𝟓
(𝑿~𝑩𝒊𝒏 𝒏 = 𝟐𝟎𝟎, 𝒑 = )
𝟐𝟖𝟗 𝟒𝟏𝟏

𝟗𝟎𝟓
Since 𝒏 = 𝟐𝟎𝟎 ≥ 𝟏𝟎𝟎 and 𝒑 = 𝟐𝟖𝟗 𝟒𝟏𝟏 < 𝟎. 𝟎𝟏, 𝑿 is approximately Poisson.
𝟗𝟎𝟓
(𝑿~𝑷𝒐𝒊𝒔𝒔𝒐𝒏 𝝀 = 𝒏𝒑 = 𝟐𝟎𝟎 ∗ 𝟐𝟖𝟗 𝟒𝟏𝟏 = .625

𝒆−𝝀 𝝀 𝟎
𝑷 𝑿≥𝟏 =𝟏−𝑷 𝑿=𝟎 =𝟏− = 𝟒𝟔. 𝟒𝟗𝟓𝟕%
𝟎!
Example
• In a certain industrial facility accidents occur infrequently. It is known
that the probability of an accident on any given day is 0.005 and
accidents are independent of each other.
1. What is the probability that in any given period of 400 days there
will be an accident in one day
2. What is the probability that there are at most three days with an
accident?
Example
• In a manufacturing process where glass products are produced,
defects or bubbles occur, occasionally rendering the piece
undesirable: for marketing. It is known that, on average, 1 in every
1000 of these items produced has one or more bubbles. What is the
probability that a random sample of 8000 will yield fewer than 7
items possessing bubbles?
Exercise

• One prominent physician claims that 70% of those with lung cancer are chain smokers. If his
assertion is correct:

• Find the probability that of 10 such patients recently admitted to a hospital, few than half are
chain smokers.

𝑋 is a binomial random variable that counts the number of lung cancer patients who
are smokers 𝑋~𝐵𝑖𝑛(𝑛 = 10, 𝑝 = 0.70)

P X < 5 = 𝑃 𝑋 ≤ 4 =.0473
Exercise

• To avoid detection at customs, a traveler places 6 narcotic tablets in a


bottle containing 9 vitamin pills that are similar in appearance. If the
customs official selects 3 of the tablets at random for analysis, what is
the probability that the traveler will be arrested for illegal possession
of narcotics.
X – hypergeometric distribution
N=15
K=6
n=3
Exercise

• It is estimated that 4000 of the 10000 voting residents of a town are


against a new sales tax. If 15 eligible voters are selected at random
and asked for their opinion, what is the probability that at most 7
favor the new tax?
4000 out of 10000 voting residents are against a new sales tax.

hence 10000 - 4000= 6000 residentds favour the new tex.

the probability that randomly selected voter favor the new tax is

p= 6000/10000 p=0.6

15 voters are selected at random n=15

𝑋 is a binomial random variable. (X~𝑩𝒊𝒏(𝒏 = 𝟏𝟓, 𝒑 = 𝟎. 𝟔𝟎))


𝟏𝟓
𝒑𝑿 𝒌 = 𝒌
. 𝟔𝒌 (𝟏−. 𝟔)𝟏𝟓−𝒌 , 𝒊𝒇 𝒌 = 𝟎, 𝟏, 𝟐, … ,15

Probability that at most 7 out of 15 randomly selected voters favor the new tex
P(X<7) = P(X=0) + P(X=1) + {(X=2)…..P(X=7) = .2131
Exercise

• Suppose that, on the average, 1 person in 1000 makes a numerical


error in preparing his or her income tax return. If 10,000 forms are
selected at random and examined, find the probability that 6,7 or 8 of
the forms contain an error.
𝑃 𝑋 ≤𝑥−1 +𝑃 𝑋 ≥𝑥 =1

𝑃 𝑋 ≥ 𝑥 = 1 − 𝑃(𝑋 ≤ 𝑥 − 1)
Example
𝑃 𝑋 ≥ 10 = 1 − 𝑃(𝑋 ≤ 9)
𝑃 𝑋 ≥ 5 = 1 − 𝑃(𝑋 ≤ 4)
𝑃 𝑋 > 10 = 1 − 𝑃 𝑋 ≤ 10
𝑃 𝑋 >5 =1−𝑃 𝑋 ≤5
𝑃 𝑋<5 =𝑃 𝑋≤4
𝑃 𝑋 < 10 = 𝑃 𝑋 ≤ 9

𝑷 𝟏𝟎 ≤ 𝑿 ≤ 𝟑𝟎 = 𝑷 𝑿 ≤ 𝟑𝟎 − 𝑷 𝑿 < 𝟏𝟎

P(at least 10) = 𝑃 𝑋 ≥ 10 = 1 − 𝑃 𝑋 ≤ 9


P(at most 10) = 𝑃 𝑋 ≤ 10
Exercise

• A manufacturer knows that on the average 20% of the electric


toasters which he makes will require repairs within 1 year after they
are sold. When 20 toasters are randomly selected, find X and Y such
that:

• The probability that at least X of them will require repairs is less than
0.5
• The probability that at least Y of them will not require repairs is
greater than 0.8.

You might also like