Integrated Business Foundation

Module 9. Decision-making Skills

Practice Problems II
Simple Linear Regression, Time Series Analysis
and Linear Programming

(For practice only; no submission is required)

Jan 2026

Please bring this problem set to each tutorial session.

1
Linear Regression
1. The following table shows the total consumer's expenditure and the total personal disposable income
in a city for the last ten years. Assuming a linear relationship, determine the linear regression
equation using the least squares method to predict consumer expenditure by the amount of personal
disposable income. Use the linear regression equation to obtain a forecast of the consumer’s
expenditure for the coming year if personal disposable income for that year was estimated to be
$65,000,000. Comment the validity of this kind of prediction.
Consumer's Personal
Expenditure Disposable Income
Year (in millions) Y (in millions) X
1 22.9 25.0
2 24.2 26.6
3 25.4 27.8
4 27.4 29.7
5 29.0 31.6
6 31.5 34.6
7 35.1 38.5
8 39.6 44.1
9 45.1 50.9
10 51.7 59.2

2. Zippy Cola is studying the effect of its latest advertising campaign. People chosen at random were
called and asked how many cans of Zippy Cola they had bought in the past week and how many
Zippy Cola advertisements they had either read or seen in the past week.

Customer 1 2 3 4 5 6 7 8
Number of ads 4 9 3 0 1 6 2 5
Number of cans purchased 12 14 7 6 3 5 6 10

(a) Assuming a linear relationship, determine the linear regression equation using the least squares
method to predict ‘number of cans purchased’ by the ‘number of ads read or seen’.
(b) Calculate the corresponding coefficient of determination and the coefficient of correlation and
interpret their meanings.

2
3. The following table shows a sampling result of the months of experience against productivity (lot
size completed per day) of a group of 10 workers in a large factory.
Worker 1 2 3 4 5 6 7 8 9 10
Experience 2 17 6 13 30 4 8 7 20 16
Productivity 4 10 7 11 14 4 6 7 15 10
Suppose that we fit a linear regression line by expressing productivity (Y) as a function of experience
(X) using the least-squares method, at the 0.05 level of significance, test whether the true slope b1 is
equal to zero or not, i.e., whether there is a linear relationship exists between experience and
productivity?

4. A production plant cost-control engineer is responsible for cost reduction. One of the costly
items in his plant is the amount of water used by the production facilities each month. He
decided to investigate water usage by collecting seventeen observations given as follows:

Water Production
Month (Monthly water usage in gallons) Amount of production (in pounds)
1 3067 7107
2 2828 6373
3 2891 6796
4 2994 9208
5 3082 14792
6 3898 14564
7 3502 11964
8 3060 13526
9 3211 12656
10 3286 14119
11 3542 16691
12 3125 14571
13 3022 13619
14 2922 14575
15 3950 14556
16 4488 18573
17 3295 15618

(a) A simple linear regression model is to be developed to predict the amount of water usage by
the amount of production, determine the corresponding linear regression equation using the
least squares method.
(b) Using the linear regression line obtained in part (a), if the target production amount is set as
20,000 pounds, provide a forecast of the amount of water usage. Comment on the suitability
of such a forecasting practice.
(c) At the 0.05 level of significance, is there enough evidence to reflect that there is a positive
linear relationship between amount of water usage and the amount of production?
(d) Compute the corresponding coefficient of determination and interpret its meaning.

3
5. The chief executive officer of a famous fast-food chain restaurant believes that the monthly
sales for each of the branch restaurants is related to the size of the working population around
each of the restaurants. He has collected data from a sample of ten branch restaurants, as
follows:
Restaurants 1 2 3 4 5 6 7 8 9 10
working population (X) 4.5 2.7 2.6 4.2 2.4 4.5 3.2 2.6 4.0 1.5
(1000)
monthly sales (Y) ($10000) 30 15 18 28 17 34 16 13 25 8

Excel output
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.953101
R Square 0.908401
Adjusted R Square 0.896951
Standard Error 2.686639
Observations 10

ANOVA
DF SS MS F Significance F
Regression 1 572.6558 572.6558 79.33688 2E-05
Residual 8 57.74422 7.218028
Total 9 630.4

Coefficients Standard Error t Stat P-value

Intercept -4.57902 2.930254 -1.56267 0.156756
Working population 7.757461 0.870928 8.907125 2E-05

(a) Use Excel to obtain the above computer output table for predicting the monthly sales on the
basis of the working [Link] using simple linear regression.
(b) Determine the sample regression equation.
(c) At the 5% level of significance, is there enough evidence to reflect that there is a linear
relationship between monthly sales and working population?
(d) At the 5% level of significance, use p-value to decide whether there is enough evidence to
reflect that there is a positive linear relationship between monthly sales and working
population.

4
6. A lecturer in Statistics wants to study the linear relationship between average number of hours spent
on mobile games per week (Hours) and GPA (GPA) over a semester of college students. A sample
of 17 students is selected at random and summarized as follows:

Hours 1 11 8 5 16 27 29 14 10 8 6 0 21 34 6 5 26
GPA 3.95 3.65 3.55 3.58 2.98 1.50 1.75 2.20 3.00 3.00 3.15 3.90 2.70 1.00 3.25 3.90 2.80

(a) Fit the linear regression model between Hours and GPA.
(b) Interpret the meaning of the regression coefficients b0 and b1 in this problem.
(c) Construct a 90% confidence interval for the population slope coefficient β1.
(d) At α = 0.01, is there evidence of a linear relationship between Hours and GPA?
(e) In this semester, Tom spends 12 hours on mobile games a week on average. Predict the GPA of
Tom.
(f) Construct a 90% confidence interval for GPA of Tom.

7. A plant manager obtained some summary information about weekly production in hundreds of units
(X) and cost per unit in dollars (Y).:

n Mean Standard Deviation

Cost per unit (Y) 102 34.5 2.0
Production (X) 102 7.0 2.5

He also calculated Sxy = −315.625.

(a) Compute SXX and SYY from the preceding table.

(b) Fit the linear regression between X and Y.
(c) Interpret the slope coefficient specifically in terms of this problem.
(d) Test the hypothesis that the cost per unit does not depend linearly on the production against a
two-sided alternative with α = 5%.
(e) Estimate the total weekly cost when the weekly production equals 1,000 units.
(f) For a particular week, the manager has budgeted a total of $38,000 to cover the weekly
production cost for producing 1,000 units. Calculate a 95% prediction interval. Do you think
that there will be any deficit? Justify your answer.

5
Answer Key

1. The regression equation is yˆ  2.1119  0.8445x

For given X=65, yˆ  2.1119  0.8445  65  57.0044 millions.
It is not appropriate to estimate Yˆ using X = 65 which is beyond the given X range (25, 59.2). An
estimate obtained using extrapolation might not be valid.

2. (a) yˆ  4.4874  0.9034 x

(b) r 2  0.4911 ; this implies that about 49.11% of the variation in Y can be explained by the variation
in X.
r = 0.7008.
This implies that there exists a rather strong positive linear relationship between X and Y.

3. H 0 : 1  0 (There is no linear relationship.)

H1 : 1  0 (There is a linear relationship.)
yˆ  3.7779  0.4083x
SSE  21.8895 ; Sb  0.0639 ; t  6.3897
1

Since t  6.3897  t8  2.306, reject H 0 .

At the 5% level of significance, there is evidence to reveal that experience and productivity are
linearly related.

4. (a) yˆ i  2,273.0880  0.0799  xi

(b) yˆ  2273.0880  0.079920,000  3,871.0880
Since x = 20,000 is beyond the range of x, it is not suitable to extrapolate to get the forecast.
(c) H 0 : 1  0; H1 : 1  0
t  3.1581  t0.05,15  1.75, reject H 0 .
At the 0.05 level of significance, there is evidence to believe that there is a positive linear
relationship between the amount of water usage and the amount of production.
(d) r 2  0.63072  0.3978
About 39.78% of the total variation in the amount of water usage can be explained by the amount
of production.

5. (b) yˆi  4.5790  7.7575  xi

(c) H 0 : 1  0; H1 : 1  0
t  8.9071  t0.025,8  2.306, reject H0.
At the 0.05 level of significance, there is evidence to reflect that there is a linear relationship
between monthly sales and working population.
(d) H 0 : 1  0 vs H 0 : 1  0
p  value  0.0000  2  0  0.05 reject H 0 .
At the 0.05 level of significance, there is evidence to reflect that positive linear relationship exists
between monthly sales and working population.

6
6. (a) Yˆ  3.9348  0.0750 X
(b) For each extra hour spent on mobile game per week, the expected GPA is estimated to drop by
0.0750.
Spending 0 hour on mobile game per week, a student is expected to get a GPA of 3.9348.
S 2
(c) SSE  S YY  XY  12.1754 
 130.24762  2.4025
S XX 1735.8824
SSE 2.4025
s   0.4002
n2 17  2
A 90% confidence interval for the population slope coefficient β1:
s 0.4002
b1  t  2  0.0750  1.753 
S XX 1735.8824
  0.0919,  0.0582

(d) H 0 :   0 vs H1 :   0
r  ρ0  0.8959  0
t obs    7.8112  2.947  t 0.005
1 r 2 1   0.8959 
2
Test statistic:
n2 17  2
 Reject H 0
There is sufficient evidence for a significant linear relationship between Hours and GPA at 1%
level of significance.
(e) Yˆ  3.9348  0.075012  3.0345
(f) A 90% prediction interval (or confidence interval) for GPA of Tom is:
1 X  x p  1 13  12
2 2
ˆ
Y  t 2 s 1    3.0345  1.753  0.4002 1  
n S XX 17 1753.8824
 2.3122, 3.7567

7. (a) S XX  s X 2  (n  1)  (2.5) 2  101  631.25

S YY  sY  (n  1)  (2.0) 2  101  404
2

S XY  315.625
(b) b1    0.5
S XX 631.25
b0  Y  b1 X  34.5  (0.5)7.0  38
Yˆ  38  0.5 X
(c) If the weekly production increases by 100 units, then the cost per unit will decrease by 0.5 dollars
on average.
(d) H 0 :  1  0 vs H 1 : 1  0
2
S XY
SSE  S YY   404  (315.625) 2 /(631.25)  246.188
S XX
s  SSE /(n  2)  246.188 /(100)  1.569

7
 b1  0  0.5
t obs    8.01  1.984  t 0.025
Test statistic: s / S XX 1.569 / 25.125
 Reject H 0
There is sufficient evidence that the cost per unit does depend linearly on the production.
(e) Yˆ  b0  b1 x  38  (0.5)(10)  33
Total weekly production cost = $33(1000) = $33,000.
(f) A 95% production interval for the cost per unit for producing 1000 units is given as:
1 X  x  2
1 7  10 2
Yˆ  t 2 s 1    33  1.984  1.569 1  
p

n S XX 102 631.25 .
 29.8498, 36.1502 
Hence we are 95% confidence that the total weekly production cost falls between $28850 and
$36150. As the interval stays below $38000, we are very confident that there will be no deficit.

8
Time Series and Forecasting
1. The sales volume (in thousands) of umbrella produced by a company for the recent nine years
(Year 9 is the current year) is listed as follows:

Year Sales (in thousands)

1 470
2 430
3 450
4 480
5 500
6 530
7 530
8 550
9 600

(a) Compute the 5-period moving average for the above series.
(b) Compute the 4-period centered moving average for the above series.

2. Gasohol (a mixture of gasoline and methyl alcohol) sales in Iowa City have been recorded over the
past ten months as shown below.

Month Jan. Feb. March April May June July Aug. Sept. Oct.
Sales 75 72 81 92 90 105 112 107 110 93
($,000)

(a) Compute the five-month moving average and the four-month centered moving average.
(b) Compute the exponentially smoothed sales with w = 0.4 and w = 0.8.
(c) Draw the time series and the two sets of exponentially smoothed values in one plot. Does there
appear to be a trend component in the time series? Which of the two exponentially smoothed
series is better in terms of smoothing?

9
3. The following is the quarterly sales for Giant International for the last three years. The sales are
reported in millions of dollars.
Quarter Year 1 Year 2 Year 3
1 65 69 70
2 46 50 55
3 98 104 108
4 136 141 150

With reference to the multiplicative model, compute the corresponding seasonal (quarterly)
indexes (to an accuracy of four decimal places) by comparing the original time series data with
their corresponding estimated values fitted by a simple linear regression line. (Utilize mean to
eliminate random effects.)

4. The following table shows the sales volume of a commodity for the last three years:
Quarter Year 1 Year 2 Year 3
1 300 350 330
2 700 720 690
3 400 390 385
4 600 610 650
With reference to the multiplicative model, compute the corresponding seasonal (quarterly)
indexes (to an accuracy of four decimal places) by comparing the original time series data with
their corresponding ratios to moving average. (Utilize mean to eliminate random effects.)

10
5. A small electronic company produces pocket calculators and records the demand monthly. For
a typical scientific calculator, the quantity of demand is as follows:

November 45
December 57
January 60
February 65

Forecast the sales in March using both the smoothing coefficients w = 0.3 and w= 0.8 and
comment on your answers.

6. The following seasonal indexes and trend linear line were computed from five years of quarterly
sales data.
Trend line: yˆ  500  30t (t  1,2,3,...,20)
Quarter Seasonal index
1 1.4
2 1.2
3 0.9
4 0.5
Forecast the sales for the next four quarters.

11
7. The net profits of a department store are reported 4 times a year, for the recent three years as the
Yt ( Yt  T  C  S  R ) values ($000) in the following table.
Year Q1 Q2 Q3 Q4
1 5 8 15 10
2 7 9 16 10
3 6 10 18 11

Assume the cyclic component (C) does not exist.

(a) Use 4-period centered moving average to eliminate the random component (R) from the time
series.
(b) Estimate Seasonal component (S) by calculating the seasonal index from the series (utilize mean
to eliminate random effects).
(c) Deseasonalize the Yt (seasonally adjusted series)
(d) Using the deseasonalized Yt to estimate the trend component (T) by least squares assuming a
linear trend.
(e) By using the least squares regression in (d), forecast the deseasonalized values for the next 4
quarters.

12
Answer Key

1. (a) 466, 478, 498, 518, 542 (b) 461.25, 477.50, 500.00, 518.75, 540.00
2. 5-month MA: 82, 88, 96, 101.2, 104.8, 105.4;
4-month centered MA: 81.875, 87.875, 95.875, 101.625, 106, 107
Exponential smoothing:
w=0.4: 75, 73.8, 76.68, 82.8080, 85.6848, 93.4109, 100.8465, 103.3079, 105.9848, 100.7909
w=0.8: 75, 72.6, 79.32, 89.464, 89.8928, 101.9786, 109.9957, 107.5991, 109.5198, 96.3040

120

110
Sales
100
w =0.4
90

w =0.8
80

70

60
n

ne
ch

ly

g
ril

pt
b

ct
ay
Ja

Au
Fe

Ju
Ap

Se

O
Ju
ar

M
M

Exponential smoothing also reflects a gradual trend of increasing sales.

3. (a) yˆ  65.2727  3.9580t
(b) Quarter 1: 0.8145; Quarter 2: 0.5715: Quarter 3: 1.1270; Quarter 4: 1.4870
The sales are low in the first two quarters but high in the last two quarters.
4. Quarter 1: 0.6666; Quarter 2: 1.3764: Quarter 3: 0.7742; Quarter 4: 1.1828
5. Using w = 0.3, the forecast for sales in March is 55.914; Using w = 0.8, the forecast for sales in
March is 63.784
6. Forecasted sales: t =21: 1582; t =22: 1392; t =23: 1071; t =24: 610
7. Four-quarter centered M.A.: 9.75, 10.125, 10.375, 10.5, 10.375, 10.375, 10.75, 11.125
Ratio to M.A.: 1.5385, 0.9877, 0.6747, 0.8571, 1.5422, 0.9639, 0.5581, 0.8989
Seasonal Index: 0.6148, 0.8757, 1.5363, 0.9732;
Deseasonalized series: 8.1327, 9.1354, 9.7638, 10.2754, 11.3857, 10.2774, 10.4147, 10.2754, 9.7592,
11.4193, 11.7166, 11.3029
Regression line: yˆ  8.8689  0.2235  t (t = 1, 2, … , 12)
Forecast: Q1 = 11.7742, Q2 = 11.9977, Q3 = 12.2212, Q4 = 12.4447

13
Linear Programming
1. A local shoe manufacturer has three factories located along the Shenzhen River (1, 2 and 3). Each
factory produces different kinds of shoes but they all emits waste that contains three types of
pollutants (1, 2 and 3) into the river. However, if the waste from each factory is processed properly,
the amount of pollutants that will be emitted into the river would be reduced. The cost to process
a ton of waste and the amount of each pollutant that can be reduced of each of the three factories
are summarized in the following table:

Amount of pollutant can be reduced (in ton)

Factory Cost to process
1 ton of waste Pollutant 1 Pollutant 2 Pollutant 3
1 $100 0.1 0.2 0.05
2 $120 0.15 0.1 0.3
3 $145 0.2 0.4 0.1

The environment protection department requires that the manufacturer to reduce at least a total of
20 tons, 30 tons and 50 tons per month of pollutant 1, 2 and 3 respectively. In algebraic form,
formulate the above problem so as to minimize the cost of reducing pollutants by the required
amounts.

2. A firm that manufactures souvenirs for tourists produces four different types of key chains. Each
type of the keychain must be sliced, polished, and assembled and the time (in minutes) required
for each of the processes are given as follows:
Type of Key Chain A B C D
Slicing 3 2 3 1
Polishing 1 1 1 1
Assembling 3 2 2 3
For a weekly production plan, the firm has available 100 labour hours of slicing time, 36 labour
hours of polishing time, and 80 labour hours of assembling time for producing the key chains.
The unit profits for each type of key chain (A, B, C, D) are $12, $10, $8, and $16 respectively.
The firm has a contract with a retailer to supply at least 400 units of type B and at least 600 units
of any combination of quantity for type A and C every week. From past experience, the firm can
sell each week as many units of type A, B and C as it can produce, but only a maximum of 200
units of type D. Formulate the linear programming model algebraically for the above situation to
determine the optimal number of each type of key chain to be produced each week to maximize
the total profit (there is no need to find the optimal solution).

14
3. Identify the maximal and minimal value of Z = 4x + 3y subject to the following constraints:
x  2 y  14
3x  y  0
x y2

4. Solve the following linear programming model graphically.

Max Z  50 X 1  20 X 2
subject to
2 X 1  4 X 2  400
100 X 1  50 X 2  8,000
X 1  60
X1 , X 2  0

5. Solve the following linear programming model graphically.

Min Z  24 X 1  28 X 2
subject to
5 X 1  4 X 2  2,000
X 1  80
X 1  X 2  300
X 2  100
X1 , X 2  0

6. Design and use Excel spreadsheet to solve the following LP problem.

To maximize Z = 2X1 + 3X2 + 5X3
subject to X1 +X2 - X3 > -5
-6X1 +7X2 - 9X3 <4
X1 +X2 + 4X3 =10
X1, X2, X3 > 0

15
7. Nicom Corporation is a camera manufacturer. Nicom produces single-lens reflex digital cameras in
her three plants located in Yangjiang, Panyu and Qingyuan. They are shipped to regional distribution
centers in Zhuhai and Guangzhou, where they are carefully packed and then delivered to the supply
houses in Shekou, Hong Kong and Macau. Excess amount of supplies will be left in the original
production sites, no stocks will be kept in any of the regional distribution centers or supply houses
at the end of each month. The shipping costs vary, as shown in the following table. Projected monthly
demands for Shekou, Hong Kong and Macau as well as the available monthly supplies of the three
plants are also given in the table.
To
From Zhuhai Guangzhou Shekou Hong Macau Supply
Kong
Yangjiang $5 $9 – – – 5,000
Panyu $4 $2 – – – 6,000
Qingyuan $7 $3 – – – 10,000
Zhuhai – – $8 $12 $3 –
Guangzhou – – $7 $14 $11 –
Demand – – 4,800 12,000 2,200

Nicom wishes to minimize the transportation costs associated with shipping sufficient cameras to
meet the demands at the three destinations. Formulate a linear programming model for the above
problem algebraically.

8. David, LaDeana, and Lydia are the sole partners and the workers in a company that produces fine
clocks. David and LaDeana are each available to work a maximum of 40 hours per week while
Lydia is available to work a maximum of 20 hours per week. The company makes 2 different types
of clocks: a grandfather clock and a wall clock. To make a clock, David (a mechanical engineer)
assembles the inside mechanical parts of the clocks while LaDeana (a woodworker) produces the
hand-carved wood casings. Lydia is responsible for taking orders and shipping the clocks. The
amount of time required for each of these tasks is shown below:
Time required
Task Grandfather clock Wall clock (hours)
(hours)
Assembles clock 6 4
Carve wood casing 8 4
Ordering and Shipping 3 3
Each grandfather clock built and shipped yields a profit of $300, while each wall clock yields a
profit of $200. The three partners want to maximize their profits.
(a) Formulate a LP model for this problem.
(b) Use both the graphical method and the Solver function to solve the LP problem.

16
 (c) Generate the Excel sensitivity report and use it to determine the allowable ranges for the unit
profits for each type of clock and the allowable range for the maximum number of hours each
partner is available to work per week.
(d) Explain why one of the shadow prices is zero.
(e) Can the shadow prices in the report be validly used to determine the effect if Lydia were to
change her maximum number of hours available to work per week from 20 to 25? If so, what
would be the increase in the total profit?
(f) In addition to revised working time for Lydia, can the shadow prices in the report be validly
used to determine the effect if David were to change his maximum number of hours available
to work per week from 40 to 35? If so, what would be the increase in the total profit?

9 . A sensitivity report generated from Excel is attached below.

(a) What is the optimal objective function value for this problem?
(b) What is the allowable range for changes in the objective coefficient for Activity 2?
(c) What is the allowable range for changes in the right-hand-side for Resource C?
(d) If the coefficient for Activity 1 in the objective function changes to $40, then what is the
change of objective function value?
(e) If the coefficient for Activity 3 in the objective function changes to $30, then what is the
objective function value?
(f) If the coefficient of Activity 1 in the objective function changes to $10, what is the objective
function value?
(g) If the right-hand side of Resource A changes to (1) 10, (2) 25, and (3) 35 respectively, what are
the resulting objective function values?
(h) If the right-hand side of Resource B changes to 10, what are the implications?
(i) If the coefficients of Activity 1 and 2 in the objective function are both increased by $10, what
can we conclude with the optimal solution? What is the revised objective function value?
(j) If the right-hand side of Resource A is increased by 4 and Resource B is increased by 20, what
is the change of the objective function value?
(k) If the right-hand side of Resource B is increased by 30, and the right-hand side of Resource C
is decreased by 10, can we estimate the change of objective function value from the Report?

17
10. Mrs Cheng has $1,000,000 available to invest in the coming year. She has two alternatives to choose
from: (1) a government bond fund that is expected to have a 5% annual rate of return, and (2) a local
high-risk stock market that is expected to have a 12% annual rate of return. To minimize risk, she
decides not to invest more than $600,000 in the stock market. To fulfil a requirement on a tax
arrangement, she needs to invest at least two times as much in the government bond as in the stock
market.
(a) Formulate the linear programming model algebraically for the above problem in order to
maximize total profit. Use the graphical method to solve this model.
(b) Use the Excel Solver to solve this model. and produce the relevant sensitivity report.

The following is the relevant sensitivity report produced by Excel:

Variable Cells
Final Reduced Objective Allowable Allowable
Cell Name Value Cost Coefficient Increase Decrease
$B$11 Bond 666666.6667 0 0.05 0.07 0.11
$C$11 Stock 333333.3333 0 0.12 1E+30 0.07

Constraints
Final Shadow Constraint Allowable Allowable
Cell Name Value Price R.H. Side Increase Decrease
$D$6 Amount available 1000000 0.073333333 1000000 800000 1000000
$D$7 Risk control 333333.3333 0 600000 1E+30 266666.6667
$D$8 Tax arrangement 0 -0.023333333 0 1000000 800000

(c) Suppose that the local stock market has experienced a sharp drop just before Mrs Cheng
engages to invest and the annual rate of return for investing into the stock market is now
estimated to be 6%, would the optimal solution obtained change and what can be said about
the effect on total profit?
(d) If Mrs Cheng is able to increase her fund to $1,500,000 to be invested into the two alternatives
by borrowing $500,000 from a bank at an annual rate of 5%, would it be worth doing so? Why
or why not? What is the new total profit that Mrs Cheng can earn?
(e) Suppose that the annual rate of return for investing into the government bond fund and the
local stock market are now revised to be 8% and 7% respectively. Without solving the linear
programming model again, can you determine whether the optimal solution given in the
sensitivity table has changed or not? Explain briefly.

18
Answer Key
1. Let X1 be the weight of waste in tons from factory 1 to be processed.
X2 be the weight of waste in tons from factory 2 to be processed.
X3 be the weight of waste in tons from factory 3 to be processed .
Minimum Cost = $100 X1 + $120 X2 + $145 X3
subject to 0.1 X1 + 0.15 X2 + 0.2 X3 > 20 tons
0.2 X1 + 0.1 X2 + 0.4 X3 > 30 tons
0.05 X1 + 0.3 X2 + 0.1 X3 > 50 tons and X1, X2, X3 ≥ 0.

2. Let A, B, C, and D be the number of units of type A, B, C, and D to be produced by the firm
respectively.
Maximize Z  12 A  10 B  8C  16 D
subject to
3A  2B  3C  D  6,000
A  B  C  D  2,160
3A  2B  2C  3D  4,800
B  400
A  C  600
D  200 and A  0, B  0, C  0, D  0

3.

Method 1: By testing all the corner points of the feasible region.

1 1
y   x7 y   x7 y  x2
2 2
y  3x y  x2 y  3x
1 1 3x  x  2
3x   x  7 x2  x7
2 2  2 x  2
 6 x   x  14  2 x  4   x  14  x  1
x2 x6 So, y  1  2  3
So, y  3( 2)  6 So, y  6  2  4
Corner point is at (2, 6) Corner point is at (6, 4) Corner point is at (–1, –3)
Z = 4x + 3y = 26 Z = 4x + 3y = 36 Z = 4x + 3y = –13
For Z = 4x + 3y,
the maximum of Z = 36 occurs at (6, 4);
the minimum of Z = –13 occurs at (–1, –3).

19
 Method 2: By graphing the objective function 4x + 3y =12 [let Z =12: (0, 4), (3, 0)]
Shift the objective function upward to obtain the maximum of Z = 36 occurs at (6, 4);
Shift the objective function downward to obtain the minimum of Z = –13 occurs at (–1, –3)

4x + 3y =12

4. X1  60; X 2  40; Z  3,800

5. X1  80; X 2  100; Z  4,720

6.
A B C D E F G
1 X1 X2 X3
2 Obj. Coefficient (Max.) 2 3 5
3
4 Coefficients Inequality Right-hand side
5 Constraints X1 X2 X3
6 1 1 1 -1 10  -5
7 2 -6 7 -9 4  4
8 3 1 1 4 10 = 10
9
10 X1 X2 X3 Maximize
11 Decision varibles 5.076923 4.923077 0 24.92307692

7. Let X YZ and X YG be the monthly amounts to be shipped from Yangjiang to Zhuhai and Guangzhou
respectively.
Let X PZ and X PG be the monthly amounts to be shipped from Panyu to Zhuhai and Guangzhou
respectively.
Let X QZ and X QG be the monthly amounts to be shipped from Qingyuan to Zhuhai and Guangzhou
respectively.
Let X ZS , X ZH and X ZM be the monthly amounts to be shipped from Zhuhai to shekou, Hong Kong
and Macau respectively.
Let X GS , X GH and X GM be the monthly amounts to be shipped from Guangzhou to shekou, Hong
Kong and Macau respectively.

20
 Minimize C = 5 X YZ +9 X YG +4 X PZ +2 X PG +7 X QZ +3 X QG +8 X ZS +12 X ZH +3 X ZM +7 X GS + 14 X GH +11
X GM
Subject to
X YZ + X YG  5000
X PZ + X PG  6000
X QZ + X QG  10000
X ZS + X GS = 4800
X ZH + X GH = 12000
X ZM + X GM = 2200
X YZ + X PZ + X QZ = X ZS + X ZH + X ZM
X YG + X PG + X QG = X GS + X GH + X GM
X YZ , X YG , X PZ , X PG , X QZ , X QG , X ZS , X GS , X ZH , X GH , X ZM , X GM  0

8. (a) Let G = number of grandfather clocks produce

W = number of wall clocks produced
Maximize Profit = $300G + $200W
subject to 6G + 4W ≤ 40 hours
8G + 4W ≤ 40 hours
3G + 3W ≤ 20 hours
and G ≥ 0, W ≥ 0.
(b) Optimal solution: 3.33 grandfather clocks and 3.33 wall clocks. Maximum profit is 1667.
(c) The allowable range for the unit profit for the grandfather clock is $200 to $400.
The allowable range for the unit profit for the wall clock is $150 to $300.
The allowable range for David’s available hours is 33.33 and above.
The allowable range for LaDeana’s available hours is 26.67 to 53.33 hours.
The allowavle range for Lydia’s available hours is 15 to 30 hours.
(d) The shadow price for David is zero because all of his available hours are not being fully
used, so an increase in his hours would not impact total profit.
(e) Yes, this increase (5 hours) is within the allowable increase (10 hours). The increase in total
profit will be ∆Z = $166.65.
(f) Percentage of Lydia’s available increase used = 50%.
Percentage of David’s allowable decrease used = 75%.
The sum is 125%, so by the 100% rule, the shadow prices may or may not be valid and hence
should not be used to determine the effect on total profit.
9. (a) $330
(b) 30 ≤ R2 ≤ 90
(c) 18 ≤ RHSc ≤ ∞
(d) $30
(e) Have to solve the model again. Possibly, the objective function will increase by an unknown
amount.
(f) Have to solve the model again.
(g) (1) Decrease by 77.8 and the resulting value = (330 – 77.8) = $252.2; (2) Increase by 38.9
and the resulting value (330 + 38.9) = $368.9; (3) The objective function value is uncertain
and we have to solve the LP again.

21
 (h) The shadow price is not valid to apply. The revised objective function value can only be
discovered by resolving the problem.
(i) By the 100% rule, the revised objective function value = $420.
(j) By the 100% rule, the objective function value will increase by $151.12.
(k) 100% rule is violated. Have to solve the LP again.

10. (a) Let X and Y be the amount to be invested in bond and stock market.

Max Z  0.05 X  0.12Y

subject to
X  Y  1,000,000
Y  600,000
X  2Y  0
X  0; Y  0

10
9
8
x+y=10
7
6 y=6
5 y=x/2
y

4
z=0.05x+0.12y
3
2
1
0
0 2 4 6 8 10
X

X  666,667; Y  333,333; Z  73,333 .3

(b)

22
 Variable Cells
Final Reduced Objective Allowable Allowable
Cell Name Value Cost Coefficient Increase Decrease
$B$11 Bond 666666.6667 0 0.05 0.07 0.11
$C$11 Stock 333333.3333 0 0.12 1E+30 0.07

Constraints
Final Shadow Constraint Allowable Allowable
Cell Name Value Price R.H. Side Increase Decrease
$D$6 Amount available 1000000 0.073333333 1000000 800000 1000000
$D$7 Risk control 333333.3333 0 600000 1E+30 266666.6667
$D$8 Tax arrangement 0 -0.023333333 0 1000000 800000

(c) The optimal solution will not change since a decrease of 0.06 is within the allowable
decrease (0.07). Total profit will decrease by $20000.

(d) An increase of available fund of $500,000 is within the allowable increase (800,000), the
shadow price of 0.07333333 would remain effective.
Thus, it is worth to do so as 0.07333333 is higher than the interest rate of 0.05.
The new total profit that Mrs Cheng can earn is: 73333.333333 + 500,000(0.02333333) =
$85,000.00
(e)
Change % of change:
Coefficient of ‘bond’ has 0.03/0.07(100%) =
increased 0.03 42.86%
Coefficient of ‘stock’ has 0.05/0.07(100%) =
decreased 0.05 71.43%

Since 42.86% + 71.43% = 114.29.73% > 100%, cannot determine whether the original
optimal solution has changed or not.

23