Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

An operations management professor wants to use the number of hours a student

studies for an operations management final exam (X) to predict the final exam score
(Y). Which of the following models should be used?

a. Simple moving average method


b. Linear regression model
c. Weighted moving average method
d. Exponential smoothing

Solution:

To study the association between an independent variable (X) and a dependent variable

(Y), linear regression model is used.


Given the following sales data, use exponential smoothing with α = 0.4 to update the
forecasts. (Assume that the actual sales for month 2 is observed only after making a
forecast for that month.)

Month 1 2 3

Actual sales 534 576

Forecast sales 560

What is the forecast sales figure (round to the nearest integer) for month 3?

a. 550
b. 555
c. 560
d. 565
e. 570

Solution:

Forecast sales in month 2 = 560 + 0.4(534 – 560) = 549.6

Forecast sales in month 3 = 549.6 + 0.4(576 – 549.6) = 560.16

Given the following sales data, use exponential smoothing with α = 0.7 to update the
forecasts. (Assume that the actual sales for month 2 is observed only after making a
forecast for that month.)

Month 1 2 3

Actual sales 534 575

Forecast sales 560

What is the forecast sales figure (round to the nearest integer) for month 3?
a. 550
b. 555
c. 560
d. 565
e. 570

Solution:

Forecast sales in month 2 = 560 + 0.7(534 – 560) = 541.8

Forecast sales in month 3 = 541.8 + 0.7(575 – 541.8) = 565.04


1. Suppose that there are 20000 seats available for the Coldplay concert. First,
20000-Q tickets are sold in advance at discounted price D before the concert
date. The rest Q tickets are reserved for walk-in customers on the day of the
concert, which are sold at full price F, where F > D. The advance tickets are always
sold out, whereas the number of walk-in customers is normally distributed with
mean 𝜇 and standard deviation 𝜎.

Which of the following statement is true?

a. To secure enough tickets for walk-in customers, the optimal protection level
(Q) must be always higher than the average number of walk-in customers (𝜇).
b. The optimal protection level (Q) depends on the mean (𝜇), but not on the
standard deviation (𝜎), of the number of walk-in customers.
c. The optimal protection level (Q) is independent of the discounted price D
because these tickets are sold to walk-in customers at full price F.
d. All else being equal, the optimal protection level (Q) increases with the
average demand from walk-in customers (𝜇).

Solution:
a. Not true, because 𝑄 = 𝜇 + 𝜎𝑧 ∗ , so Q may be lower than 𝜇 if 𝑧 ∗ is negative (or
equivalent, if the critical fractile is less than 0.5)
b. Not true, because the optimal protection level depends on both the mean and
the standard deviation of the number of walk-in customers
c. Not true, because the discounted price D affects the critical fractile, and thus it
also affects 𝑧 ∗ .
1. Suppose there are 100 seats on a flight from Hong Kong to Seoul. The number of
no-shows is normally distributed with mean 20 and standard deviation 10. The
price of an airline ticket is p. The cost per bumped customer is c. Assume c > p.
Choose all of the statements that are true.
a. The airline should sell less than 120 tickets.
b. If the price of the ticket increases, the optimal overbooking level decreases.
c. If the cost for bumped customers by half, then the optimal overbooking
level decreases.
d. If the number of no-shows is indeed normally distributed with mean 20 and
standard deviation 5, the optimal overbooking level is higher than that in
the original scenario.

Solution:
C means Co, p means Cu. When c > p, critical fractile < 0.5, the number of tickets
overbooked = 20 + z(<0.5)*10, z(<0.5) is a negative number, the overbooking ticket is less
than 20. The airline is selling 100 tickets + overbooking level (<20). The airline should
sell less than 120 tickets.

If p increases, Cu increases, critical fractile increases, z-value increases, the optimal


overbooking level increases.

If c is halved, Co decreases, critical fractile increases, z-value increases, the optimal


overbooking level increases.

If the mean of no-show is 20, with a standard deviation of 5, the optimal overbooking
level = 20 + z(<0.5)*5, with a smaller standard deviation, the overbooking level is higher.

2. Suppose there are 100 seats on a flight from Hong Kong to Seoul. The number of
no-shows is normally distributed with mean 20 and standard deviation 5. The price
of an airline ticket is p. The cost per bumped customer is c. Assume c > p. Choose
all of the statements that are true.
a. The airline should sell more than 120 tickets.
b. If the price of the ticket increases, the optimal overbooking level increases.
c. If the cost for bumped customers by half, then the optimal overbooking
level decreases.
d. If the number of no-shows is indeed normally distributed with mean 20 and
standard deviation 10, the optimal overbooking level is lower than that in
the original scenario.
Solution:
C means Co, p means Cu. When c > p, critical fractile < 0.5, the number of tickets
overbooked = 20 + z(<0.5)*5, z(<0.5) is a negative number, the overbooking ticket is less
than 20. The airline is selling 100 tickets + overbooking level (<20). The airline should
sell less
than 120 tickets.

If p increases, Cu increases, critical fractile increases, z-value increases, the optimal


overbooking level increases.

If c is halved, Co decreases, critical fractile increases, z-value increases, the optimal


overbooking level increases.

If the mean of no-show is 20, with a standard deviation of 10, the optimal overbooking
level = 20 + z(<0.5)*10, with a larger standard deviation, the overbooking level is lower.

3. Suppose there are 100 seats on a flight from Hong Kong to Seoul. The number of
no-shows is normally distributed with mean 20 and standard deviation 10. The
price of an airline ticket is p. The cost per bumped customer is c. Assume c > p.
Choose all of the statements that are true.
a. The airline should sell more than 120 tickets.
b. If the price of the ticket increases, the optimal overbooking level increases.
c. If the cost for bumped customers by half, then the optimal overbooking
level decreases.
d. If the number of no-shows is indeed normally distributed with mean 20 and
standard deviation 5, the optimal overbooking level is higher than that in
the original scenario.

Solution:
C means Co, p means Cu. When c > p, critical fractile < 0.5, the number of tickets
overbooked = 20 + z(<0.5)*10, z(<0.5) is a negative number, the overbooking ticket is less
than 20. The airline is selling 100 tickets + overbooking level (<20). The airline should
sell less than 120 tickets.

If p increases, Cu increases, critical fractile increases, z-value increases, the optimal


overbooking level increases.

If c is halved, Co decreases, critical fractile increases, z-value increases, the optimal


overbooking level increases.
If the mean of no-show is 20, with a standard deviation of 5, the optimal overbooking
level = 20 + z(<0.5)*5, with a smaller standard deviation, the overbooking level is higher.

You might also like