Chapter 4.

Part 4

October 4, 2021

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 1/7

 Chapter 4. Part 4

October 4, 2021

1 4.7 Approximation of Binomial, Poisson distributions

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 2/7

Normal approximation to the Binomial distribution

Theorem
If X = B(n, p) is a binomial random variable, then

X − np
Z=p
np(1 − p)

is approximated standard normal.

We can approximate
!
a + 0.5 − np
P(X ≤ a) = P(X ≤ a + 0.5) ≈ P Z ≤ p
np(1 − p)
!
a − 0.5 − np
P(X ≥ a) = P(X ≥ a − 0.5) ≈ P Z ≥ p
np(1 − p)

The approximation is good for np > 5 and n(1 − p) > 5.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 3/7

Normal approximation to the Binomial distribution

Theorem
If X = B(n, p) is a binomial random variable, then

X − np
Z=p
np(1 − p)

is approximated standard normal.

We can approximate
!
a + 0.5 − np
P(X ≤ a) = P(X ≤ a + 0.5) ≈ P Z ≤ p
np(1 − p)
!
a − 0.5 − np
P(X ≥ a) = P(X ≥ a − 0.5) ≈ P Z ≥ p
np(1 − p)

The approximation is good for np > 5 and n(1 − p) > 5.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 3/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that more than 30 chips are defective.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 4/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that more than 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(X ≥ 31) by the normal
distribution.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 4/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that more than 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(X ≥ 31) by the normal
distribution.
30.5 − 800 ∗ 0.03
P(X ≥ 31) = P(X ≥ 30.5) ≈ P(Z ≥ √ )
800 ∗ 0.03 ∗ 0.97
=P(Z > 1.347) = 1 − Φ(1.347) = 1 − 0.911 = 0.089.

In fact,
800
!
X 800
P(X ≥ 31) = 0.03x 0.97800−x = 0.09254
x =31
x

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 4/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that more than 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(X ≥ 31) by the normal
distribution.
30.5 − 800 ∗ 0.03
P(X ≥ 31) = P(X ≥ 30.5) ≈ P(Z ≥ √ )
800 ∗ 0.03 ∗ 0.97
=P(Z > 1.347) = 1 − Φ(1.347) = 1 − 0.911 = 0.089.

In fact,
800
!
X 800
P(X ≥ 31) = 0.03x 0.97800−x = 0.09254
x =31
x

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 4/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that between 20 and 30 chips are defective.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 5/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that between 20 and 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(21 ≤ X ≤ 29) by normal
distribution.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 5/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that between 20 and 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(21 ≤ X ≤ 29) by normal
distribution.

P(21 ≤ X ≤ 29) = P(20.5 ≤ X ≤ 29.5)

20.5 − 800 ∗ 0.03 29.5 − 800 ∗ 0.03
 
≈P √ ≤Z ≤ √
800 ∗ 0.03 ∗ 0.97 800 ∗ 0.03 ∗ 0.97
=P(−0.725 ≤ Z ≤ 1.14) = Φ(1.14) − Φ(−0.725) = 0.873 − 0.234 = 0.639

In fact,
29
!
X 800
P(21 ≤ X ≤ 29) = 0.03x 0.97800−x = 0.632
x =21
x

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 5/7

Example
The manufacturing of semiconductor chips produces 3% defective chips.
Assume the chips are independent and that a lot contains 800 chips.
Approximate the probability that between 20 and 30 chips are defective.
Answer: Let X be the number of defective chips from the lot. Then
X = B(800, 0.03). We want to approximate P(21 ≤ X ≤ 29) by normal
distribution.

P(21 ≤ X ≤ 29) = P(20.5 ≤ X ≤ 29.5)

20.5 − 800 ∗ 0.03 29.5 − 800 ∗ 0.03
 
≈P √ ≤Z ≤ √
800 ∗ 0.03 ∗ 0.97 800 ∗ 0.03 ∗ 0.97
=P(−0.725 ≤ Z ≤ 1.14) = Φ(1.14) − Φ(−0.725) = 0.873 − 0.234 = 0.639

In fact,
29
!
X 800
P(21 ≤ X ≤ 29) = 0.03x 0.97800−x = 0.632
x =21
x

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 5/7

Normal approximation to the Poisson distribution

Theorem
If X is a Poisson random variable with parameter λ, then
X −λ
Z= √
λ
is approximately a standard normal random variable.
We can approximate:
a + 0.5 − λ
 
P(X ≤ a) = P(X ≤ a + 0.5) ≈ P Z ≤ √
λ
a − 0.5 − λ
 
P(X ≥ a) = P(X ≥ a − 0.5) ≈ P Z ≥ √
λ

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 6/7

Normal approximation to the Poisson distribution

Theorem
If X is a Poisson random variable with parameter λ, then
X −λ
Z= √
λ
is approximately a standard normal random variable.
We can approximate:
a + 0.5 − λ
 
P(X ≤ a) = P(X ≤ a + 0.5) ≈ P Z ≤ √
λ
a − 0.5 − λ
 
P(X ≥ a) = P(X ≥ a − 0.5) ≈ P Z ≥ √
λ
The approximation is good for λ > 5.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 6/7

Normal approximation to the Poisson distribution

Theorem
If X is a Poisson random variable with parameter λ, then
X −λ
Z= √
λ
is approximately a standard normal random variable.
We can approximate:
a + 0.5 − λ
 
P(X ≤ a) = P(X ≤ a + 0.5) ≈ P Z ≤ √
λ
a − 0.5 − λ
 
P(X ≥ a) = P(X ≥ a − 0.5) ≈ P Z ≥ √
λ
The approximation is good for λ > 5.

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 6/7

Example
The number of customers that arrive at a fast-food business during a
one-hour period is known to be Poisson distributed with a mean equal to
9.6. What is the probability that more than 10 customers will arrive in a
one-hour period?

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 7/7

Example
The number of customers that arrive at a fast-food business during a
one-hour period is known to be Poisson distributed with a mean equal to
9.6. What is the probability that more than 10 customers will arrive in a
one-hour period?
Answer. Let X be the number of customers that arrive. X = has a
Poisson distribution with parameter λ = 9.6. We want to approximate
P(X ≥ 11).

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 7/7

Example
The number of customers that arrive at a fast-food business during a
one-hour period is known to be Poisson distributed with a mean equal to
9.6. What is the probability that more than 10 customers will arrive in a
one-hour period?
Answer. Let X be the number of customers that arrive. X = has a
Poisson distribution with parameter λ = 9.6. We want to approximate
P(X ≥ 11).

10.5 − 9.6
 
P(X ≥ 11) = P(X ≥ 10.5) ≈ P Z ≥ √
9.6
=P(Z ≥ 0.29) = 1 − Φ(0.29) = 1 − 0.614 = 0.386

In fact,
∞
X e −9.6 9.6x
P(X ≥ 11) = = 0.367
x =11
x!

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 7/7

Example
The number of customers that arrive at a fast-food business during a
one-hour period is known to be Poisson distributed with a mean equal to
9.6. What is the probability that more than 10 customers will arrive in a
one-hour period?
Answer. Let X be the number of customers that arrive. X = has a
Poisson distribution with parameter λ = 9.6. We want to approximate
P(X ≥ 11).

10.5 − 9.6
 
P(X ≥ 11) = P(X ≥ 10.5) ≈ P Z ≥ √
9.6
=P(Z ≥ 0.29) = 1 − Φ(0.29) = 1 − 0.614 = 0.386

In fact,
∞
X e −9.6 9.6x
P(X ≥ 11) = = 0.367
x =11
x!

ThienNV (FPTU) Chapter 4. Part 4 October 4, 2021 7/7