Exercises from chapter 4

4.1 The table shows the number of units of blood type A that the hospital
Woodlawn has been used in the last 6 weeks.

Week of Units Employees

August 31 360

September 7 389

September 14 410

September 21 381

September 28 3 weeks 368

October 5 374

a. forecast the demand for the week of October 12 with a

3-week moving average.

Sep 21 = 381 +

Sep28 = 368

Oct 5 = 374

1,123/3374.33 moving average week of October 12.

Use a three-week weighted moving average with weights of
1. 3 and 6 using 6 for the recent week. Forecast the demand for the
week of October 12.

August 31 360

September 7 389

September 14 410

September 21 381

September 28 3 weeks 368

October 5 374

SEP 21410(0.6)+389(0.3)+360(0.1)= 246+116.7+36= 398.7/1= 398.7

SEP 28381(0.6)+410(0.3)+389(0.1)= 228.6+123+38.9= 390.5/1 =390.5

OCT 5 368(0.6)+381(0.3)+410(0.1)= 220.8+114.3+41= 376.1/1= 376.1

OCT 12374(0.6)+368(0.3)+381(0.1)= 224.4+110.4+38.1= 372.9/1=372.9

c. Calculate the forecast for the week of October 12 by applying
exponential smoothing with a forecast of 360 for August 31 and
a = .2

WEEKS FORECAST UNITS

August 31 360 360

September 7 389 360 360+0.2 (360-360)=360

September 14 410 365.08 365.8

September 21 381 374.64 365.08 + 0.2(410 - 365.08) = 374.64

September 28 368 375.91 374.64 + 0.2(381 - 374.64) = 375.91

October 5 374 374.33 375.91 + 0.2(368 - 375.91) = 374.33

October 12 374.26 374.33 + 0.2(374 - 374.33) = 374.26

Demand forecast for the week of October 12 = 374.26

4.2 Year 1 2 3 4 5 6 7 8 9 10 11

Demand 7 9 5 9 13 8 12 13 9 11 7

A) Graph the previous data and observe any trends, cycles, or

random variations?

B) Starting from year 4 and until year 12, forecast the demand using
3-year moving averages. Graph forecasts on the same graph as the
original data.

Year 1 2 3 4 5 6 7 8 9 10 11

Demand 7 9 5 9 13 8 12 13 9 11 7

9 + 13 + 8/3 = 10

13+8+12/3=11

8+12+13/3= 11

12+13+9/3=11.3

13 + 9 + 11 / 3 = 11

9 + 11 + 7 / 3 = 9

The forecast for year 12 has a demand of 9

 14
12
10
8 year
6 demand
4
2
0
1 3 5 7 9 11 13

C) Starting from year 4 to year 12, forecast the demand

using a 3-year moving average, comparisons of .1 .3 .6 using 6
for the recent year GRAPH your forecast on the same graph.

Year demand Weighted moving average 3 years

1 7

2 9

3 5

4 9 9(6)+5(3)+9(1) /10= 7.8

5 13 13(6)+9(3)+5(1) /10= 11

6 8 8(6)+13(3)+9(1) /10=9.6

7 12 12(6)+8(3)+13(1) /10=10.9

8 13 13(6) + 12(3) + 8(1) / 10 = 12.2

9 9 9(6)+13(3)+12(1)/10= 10.5

10 11 11(6)+9(3)+13(1)/10=10.6

11 7 7(6)+11(3)+9(1)/10=8.4

12
 14
12
10
8 year
6 demand
4
2
0
1 2 3 4 5 6 7 8 9 10 11 12

D) When observing the forecast against the original data, which do you consider that
provides better results.

R= Provides better results in year 6 with an average forecast of 9.6.

in comparison with the actual demand 8.
4.3 Return to problem 4.2. Develop a forecast for years 2 to 12
through exponential smoothing with alpha=.4 and a forecast for the 1st of June.
Graph your new forecast along with the actual data and a forecast
intuitive, based on your visual inspection. Which forecast is better?

R=

Demand Forecast

1 7 6

2 9 6.4 6 + 0.4 (7 - 6) = 6.4

3 5 7.44 6.4 + 0.4 (9 - 6.4) = 7.6

4 9 6.4640 7.44 + 0.4 (5 - 7.44) = 6.4640

5 13 7.4784 6.4640+0.4 (9-6.4640)=7.4784

6 8 9.6870 7.4784 + 0.4 (13 - 7.4784) = 9.6870

7 12 9.0122 9.0122

8 13 10.2073 9.0122 + 0.4 (12 - 9.0122) = 10.2073

9 9 11.3244 10.2073 + 0.4 (13 - 10.2073) = 11.3244

10 11 10.3946 11.3244 + 0.4 (9 - 11.3244) = 10.3946

11 7 10.6368 10.3946 + 0.4 (11 - 10.3946) = 10.6368

12 9 9.1821 10.6368 + 0.4 (7 - 10.6368) = 9.1821

The best forecast is for year 9 with 11.3244.

A check processing center uses smoothing.
exponential to forecast the number of incoming checks per month. The
the number of checks received in June was 40 million, while the
The forecast was 42 million. A smoothing constant of .2 was used.

R=

What is the forecast for July?

New forecast for July = 42+.2 (40-42)=41.6 million checks.

If the center received 45 million in July, what will the forecast be for August?

41.6 + 0.2 (45 - 41.6) = 42.28 million checks.

c. For what reason might this forecasting method be inappropriate for this
situation?

R/ Because the number of checks received could increase in the following

or to maintain.
Carbondale Hospital is considering purchasing a new ambulance.
The decision will depend, in part, on the number of miles that will have to be driven.
next year. The miles traveled during the previous 5 years are the
following:

Year Miles 2-year moving average

1 3,000

2 4,000

3 3,400 3500

4 3,800 4,000 + 3,400 / 2 = 3,700

5 3,700 3,400 + 3,800 / 2 = 3,600

6 3,800 + 3,700 / 2 = 3,750

a. forecast the number of miles for next year with an average

2-year-old mobile.

R= 3,750

b. Find the MAD for your forecast from item a.

Year real demand forecast absolute deviation

1 3,000

2 4,000

3 3,400 3,500 -100

 4 3,800 3,700 -100

5 3,700 3,600 -100

6 3,750 300/3=100

c. Use a weighted moving average of 2 years with weights of .4 and .6

to forecast the number of miles next year (.6 the weight of the year plus
recent) What is the MAD of this forecast?

Year miles MAD forecast

1 3,000

2 4000

3 3,400 3,600 200 4000(0.6)+3000(0.4)/1=3,600

4 3,800 3,640 160 3,640

5 3,700 3,640 60 3,800(0.6)+3,400(0.4)/1=3640

6 3,740 420/3=1403700(0.6)+3,800(0.4)/1=3740

d. calculate the forecast for year 6 using exponential smoothing, a

initial forecast for year 1 of 3,000 miles and a= .5

Year miles forecast

1 3000 3000

2 4000 3000 3000+.5(3000-3000)=3000

3 3400 3500 3000 + 0.5(4000 - 3000) = 3500

4 3800 3450 3500 + 0.5 (3400 - 3500) = 3450

5 3700 3625 3450 + 0.5 (3800 - 3450) = 3625

6 3662.5 3625 + 0.5 (3700 - 3625) = 3662.5

4.6 Sales at Bettern Inc. were as follows:

Me Sales

January 20

February 21

March 15

April 14

May 13

June 16

July 17

August 18

September 20

October 20

November 21

December 23

25
20
15 month
10 sales
5
0
1 3 5 7 9 11 13
months
a.

MES SALES 3 MONTH MOVING AVERAGE

January 20

February 21

March 15

April 14 20+21+15=56/3=18.66

May 13 21+15+14=50/3=16.66

June 16 15 + 14 + 13 = 42 / 3 = 14

Julius 17 14 + 13 + 16 = 43 / 3 = 14.33

August 17 13 + 16 + 17 = 46 / 3 = 15.33

September 18 16+17+18=51/3=17

October 20 17 + 18 + 20 = 55 / 3 = 18.33

November 21 18 + 20 + 20 = 58 / 3 = 19.33

December 23 20+20+21=61/20.33
Potential smoothing

a = .3 and a forecast of September 18

MES SALES FORECASTS

January 20

February 21

March 15

April 14

May 13

June 16

July 17

August 18

September 20 18

October 20 18.6 18.6

November 21 19.02 18.6+0.3 (20-18.6)=19.02

December 23 19.61 19.02+0.3 (21-19.02)=19.61

JANUARY 20.63 19.61+0.3 (23-19.61)=20.63

c. With the given data, what method would allow you to prepare the sales forecast?
for next March? The potential softening.
Doug Moodie is the president of Garden Product Limited. For the last 5
years have lost its vice presidents of marketing and operations to give it
sales forecasts. Actual sales and forecasts are presented in the table.
According to MAD, which of the two vice presidents performed better?
forecast?

YEAR SALES VP/MARKETING MADVP/OPERATIONS MAD

1 167.325 170,000 2.675 160,000 7.33

2 175.362 170,000 5.362 165,000 10.3620

3 172.536 180,000 4.464 170,000 2.54

4 156.732 180,000 23.27 175,000 18.27

5 176.325 165,000 11.33 165,000 11.33

47.1010 divided by 5 equals 9.420249.832 divided by 5 equals 9.9664

The vice president who presented the best forecast is the one from marketing with a margin.
of error de9.4202

4.8 The daily high temperatures in the city of Houston during the last week
were the following 93, 94, 9396, 98, 90 (yesterday).

a.

DAYS TEMPERATURE 3-DAY MOVING AVERAGE

L 93

M 94

M 93

J 96 a: 93 + 94 + 93 = 280 / 3 = 93.33

V 88 94 + 93 + 96 = 283 / 3 = 94.33

S 90 93 + 96 + 88 = 277 / 3 = 92.33

D 96 + 88 + 90 = 274 / 3 = 91.33
B/ forecast the high temperature for today using a 2-day moving average

Days temperature 2-day moving average

L 93

M 94

M 93 93 + 94 / 2 = 93.5

J 95 94+93/2=93.5

V 96 93 + 95 / 2 = 94

S 88 95 + 98 / 2 = 96.5

D 90 98 + 88 / 2 = 93

88+90=89 temperature

c. calculate the mean absolute deviation based on a 2-day moving average.

DAYS TEMPERATURE 2-DAY MOVING AVERAGE MAD

L 93

M 94

M 93 93 + 94 / 2 = 93.5 93-93.5= 0.5

J 95 94+93/2=93.5 96-93.5= 2.5

V 96 93 + 96 / 2 = 94.5 88-94.5= 6.5

S 88 96 + 88 / 2 = 92 90-92= 2

D 90 88+90=89 2.875
d. calculate the mean squared error for a 2-day moving average.

DAYS TEMPERATURE 2-Day Moving Average QUADRATIC ERROR

L 93

M 94

M 93 93.5 0.5*0.5=0.25

J 96 93.5 2.5*2.5=6.25

V 88 94.5 6.5 * 6.5 = 42.25

S 90 92 2*2=4

D 89 52.75/4=13.1875

e. calculate the mean absolute percentage error for the 2-day moving average.

DAYS TEMPERATURE 2-Day Moving Average PERCENTAGE ERROR

L 93

M 94

M 93 93.5 0.5376

J 96 93.5 100(2.5/96)=2.6042

V 88 94.5 100(6.5/88)=7.3864

S 90 92 100(2/90)= 2.22

D 89 12.7482/4= 3.1871
4.9P.D. uses an x63 chip in some of its computers. The prices of the chip
during the last 12 months.

MES PRICE PER CHIP 2-Month Moving Average

January $1.80

February 1.67

March 1.70 1.80 + 1.67 = 3.47 / 2 = 1.735

April 1.85 1.67 + 1.70 = 3.37 / 2 = 1.69

May 1.90 1.70 + 1.85 = 3.55 / 2 = 1.78

June 1.87 1.85 + 1.90 = 3.75 / 2 = 1.875

July 1.80 1.90+1.87=3.77/2=1.89

August 1.83 1.87 + 1.80 = 3.67 / 2 = 1.84

September 1.70 1.80 + 1.83 = 3.63 / 2 = 1.82

October 1.65 1.83 + 1.70 = 3.53 / 2 = 1.77

November 1.70 1.70 + 1.65 = 3.35 / 2 = 1.68

December 1.75 1.65 + 1.70 = 3.35 / 2 = 1.68

1.70 + 1.75 = 3.45 / 2 = 1.73

2
1.9 price per chip
1.8
1.7 moving average
1.6 2months
1.5
1 3 5 7 9 11 13
average
b. Use a 3-month moving average and add it to the graph created in task a.

My price per chip 3-month moving average

January $1.80

February 1.67

March 1.70

April 1.85 1.80 + 1.67 + 1.70 / 3 = 1.7233

May 1.90 1.67 + 1.70 + 1.85 / 3 = 1.74

June 1.87 1.70 + 1.85 + 1.90 / 3 = 1.8166

July 1.80 1.85 + 1.90 + 1.87 / 3 = 1.8733

August 1.83 1.90 + 1.87 + 1.80 / 3 = 1.8566

September 1.70 1.8333

October 1.65 1.80 + 1.83 + 1.70 / 3 = 1.7766

November 1.70 1.83 + 1.70 + 1.65 / 3 = 1.7266

December 1.75 1.70 + 1.65 + 1.70 / 3 = 1.6833

1.65 + 1.70 + 1.75 / 3 = 1.7

2
1.9
price per chip
1.8
1.7 moving average
1.6 2 months
1.5 moving average
1 3 5 7 9 11 13 3 months
average
c. which is better (using the mean absolute deviation) the average of 2 months or the
average of 3 months?

R= using the mean absolute deviation, I would think it is the best for calculating.
forecasts.

d. calculate the forecast for each month with exponential smoothing and a
initial forecast for January of $1.80 to a=0.3 and finally to a=5 according to MAD which is a
better?

MES PRICE FORECAST EXPONENTIAL SMOOTHING.

January 1.80 $1.80

February 1.67 1.84 1.80 + 0.3(1.80 - 1.80) = 1.80

March 1.70 1.79 1.80 + 0.3 (1.80 - 1.67) = 1.84

April 1.85 1.763 1.84 + 0.3 (1.67 - 1.84) = 1.79

May 1.90 1.79 1.79 + 0.3 (1.70 - 1.79) = 1.763

June 1.87 1.823 1.763+0.3 (1.85-1.763)=1.79

July 1.80 1.8371 1.823

August 1.83 1.8260 1.823+0.3 (1.87-1.823)1.8371

September 1.70 1.8272 1.8371 + 0.3 (1.80 - 1.8371) = 1.8260

October 1.65 1.7890 1.8260 + 0.3 (1.83 - 1.8260) = 1.8272

November 1.70 1.7473 1.8272+0.3 (1.70-1.8272)=1.7890

December 1.75 1.7331 1.7890 + 0.3 (1.65 - 1.7890) = 1.7473

1.7473 + 0.3 (1.70 - 1.7473) = 1.7331

4.10 the data collected in the annual registrations for a Six seminar
Sigma in Quality, Collage, is shown in the following table.

a.

YEAR REGISTRATIONS (thousands) 3-MONTH MOVING AVERAGE

1 4

2 6

3 4

4 5 4 + 6 + 4 = 14 / 3 = 4.66

5 10 6 + 4 + 5 = 15 / 3 = 5

6 8 4+5+10=19/3=6.33

7 7 5+10+8=23/3=7.66

8 9 10 + 8 + 7 = 25 / 3 = 8.33

9 12 8 + 7 + 9 = 24 / 3 = 8

10 14 7+9+12=28/3=9.33

11 15 9+12+14=35/3=11.66

12 12 + 14 + 15 = 41 / 3 = 13.66
b.

Estimate the demand again for the years 4 to 12 using a moving average.
weighted where the enrollment of the most recent year has a weight of 2 in the
two more years one peso of 1.

YEAR registrations

1 4

2 6

3 4

4 5 4(2) + 6(1) + 4(1) = 18 / 4 = 4.5

5 10 5(2)+4(1)+6(1)=20/4=5

6 8 10(2) + 5(1) + 4(1) = 29/4 = 7.25

7 7 8(2) + 10(1) + 5(1) = 31/4 = 7.75

8 9 7(2)+8(1)+10(1)=32/4=8

9 12 9(2)+7(1)+8(1)=33/4=8.25

10 14 12(2)+9(1)+7(1)=40/4=10

11 15 14(2) + 12(1) + 9(1) = 49/4 = 12.25

12 15(2)+14(1)+12(1)=56/4=14

c. Graph the original data and the forecasts? Which of the methods seems
better?
 20
15 registrations
10
moving average
5 3 months
0 average
1 3 5 7 9 11 weighted
years

The 3-month moving average is the one that is closest to the enrollments.

4.11 Use exponential smoothing with smoothing constants of 0.3 for

forecast the registrations for the seminar on problem 4.10 to begin the
Procedure assumes that the forecast for year 1 was an enrollment of 5,000.
people

Year registrations (thousands) exponential smoothing

1 4 5.0

2 6 4.7 5.0 + 0.3 (4 - 5.0) = 4.7

3 4 5.09 4.7 + 0.3 (6 - 4.7) = 5.09

4 5 4.763 5.09+0.3 (4-5.09)=4.763

5 10 4.8341 4.763 + 0.3 (5 - 4.763) = 4.8341

6 8 6.3839 4.8341 + 0.3 (10 - 4.8341) = 6.3839

7 7 6.8687 6.3839 + 0.3 (8 - 6.3839) = 6.8687

8 9 6.9091 6.9091

9 12 7.5364 6.9091+0.3 (9-6.9091)=7.5364

10 14 8.8755 7.5364 + 0.3 (12 - 7.5364) = 8.8755

11 15 10.4129 8.8755 + 0.3 (14 - 8.8755) = 10.4129

12 11.7890 10.4129 + 0.3 (15 - 10.4129) = 11.7890

4.12 In problems 4.10 and 4.11, three forecasts were developed.
registrations for the seminar that are mobile forecasting for 3 months, weighted and
exponential smoothing. With MAD as a criterion, which of the three methods of
Forecasts are better, can you explain your answer?

Prom. Prom. Fabric softener

Month Mobile Registration MobileMAD Weighted MAD Exponential MAD

1 4 5 1.00
2 6 4.7 1.30
3 4 5.09 1.09
4 5 4.67 0.32 4.5 0.50 4.8 0.20
5 10 5 5.00 5 5.00 4.8 5.20
6 8 6.33 1.67 7.25 0.75 6.3 1.70
7 7 7.67 1.67 7.75 0.75 6.9 0.10
8 9 8.33 0.67 8 1.00 6.9 2.10
9 12 8 4.00 8.25 3.75 6.9 5.10
10 14 9.33 4.67 10 4.00 8.5 5.50
11 15 11.67 3.33 12.25 2.75 10.1 4.90

20.33 18.50 28.19

Deviations 2.54 2.31 2.56

The best forecasting method is the Weighted Moving Average... because its MAD is smaller 2.31
than the other two
Forecast
4.13 As shown in the following table, the demand for surgery for the
heart transplants at Washington General Hospital have increased significantly
stable over the last few years.

Transplant
A. of
Fabric softener
Year Heart Softener = .6 =0.9
1 45 41 41
2 50 43 = 41 + 0.6(45 - 41) 45
3 52 47 = 43.4 + 0.6(50 - 43.4) 49
4 56 50 = 47.36 + 6(52 - 47 - 36) 52
5 58 54 = 50.14 + .6(56.50.14) 56
6 ? 56=53.65+0.6(58-53-65) 58

Transplant
B. of
Moving Average
Year Heart 3 years
1 45
2 50
3 52
4 56 (45+50+52)/3=49
5 58 (50+52+56)/3=53
6 ? (52+56+58)/3=55

C. Transplant of Projection of
Year X Heart And X2 XY Trend
1 45 1 45 2.8 + 13.23 (1) = 16.04
2 50 4 100 2.8 + 13.23 (2) = 29.27
3 52 9 156 2.8 + 13.23 (3) = 36.88
4 56 16 224 2.8 + 13.23 (4) = 50.39
5 58 25 290 2.8 + 13.23 (5) = 68.96
 6 55 36 330 2.8 + 13.23 (6)=82.19
Ex= 21 Ey=261 Ex2=91 Exy=1,145

X = 21/6 = 3.5

Y = 261/6 = 43.5

b = 1,145 - (6) (3.5) (43.5) 231.5 13.23

91-(6) (3.5"2) 17.5

a = 43.5 - 13.23 (3.5) 2.18

The least squares equation for the trend is Y = 2.81 plus 13.23x

4.14 go back to problem 4.13 based on the MAD criterion. Which of the 4
is forecasting better?

AVERAGE TRANSPLANT Conditioner SAV.

YEAR HEART MOBILE MAD =6 MAD =9 MAD

1 45 41 4.00 41 4.00

2 50 43 7.00 45 5.00

3 52 47 5.00 49 3.00

4 56 49 7.00 50 6.00 52 4.00

5 58 52.67 5.00 54 4.00 56 2.00

MAD 6.2 5.20 3.60

Trend Projection 0.64 MAD is the best because it is the smallest.

than the other forecasts

Exponential Softener .6
MAD= 5.2

Exponential Softener .9
MAD= 3.6
MAD Moving Average=6.2

YEAR TRANS. OF HEART X2 XY TREND MAD

1 45 1 45 45.8 0.80

2 50 4 100 49 1.00

3 52 9 156 52.2 0.20

4 56 16 224.00 55.4 0.60

5 58 25 290.00 58.6 0.60

/5=0.64

X= 15/5
=3

Y = 261/5 = 52.2

b = 815 - (5) (3) (52.2) 32 3.2

55-(5) (3"2) 10

a=52.2 - 3.2 (3) 42.6

The least squares equation for the trend is Y = 42.6 + 3.2x

4.15 Take up the solved problem 4.1 from page 136. Use a 3-year mobile for.
forecast the sales of Volkswagen Beetle in Nevada during 2003.
Year sales 3-year moving average

1998 450

1999 495

2000 518

2001 563 450+495+518/3=487.66

2002 584 495 + 518 + 563 / 3 = 525.33

2003 ¿ 518+563+584/3=555 sales are forecasted

The sales of 2003 according to the 3-year moving average forecast would be 555.

4.16 Revisit the solved problem from 4.1 using the projection method.
trend forecasted Volkswagen Beetle sales in Nevada 2003.

YEAR X SALES X2 XY

1998 1 450 1 450

1999 2 495 4 990

2000 3 518 9 1554

2001 4 563 16 2252

2002 5 584 25 2920

EX= 15 EY = 2,610 EX2 = 55 8,166

X = 15/5 = 3

Y = 2,610 / 5 = 522

B= 8,166-(5) (3) (522) 336 33.6

55 - (5) (3 '' 2) 10

A = 522 - 33.6 (3) 421.2

The least squares equation for the trend is Y = 421.2 + 33.6

4.17 return to the solved problem 4.1 with smoothing constants of .6 .9 and
forecast Volkswagen sales. What effect does the smoothing constant have?
use MAD to determine which of the three smoothing constants (3.6.9) gives the
most accurate forecast.
Year units forecast 0.3 MAD 0.6 MAD 0.9 MAD

1998 450 410 40 410 40 410 40

1999 495 422 73 434 61 446 49

2000 518 443.9 74.1 470.6 47.4 490.1 27.9

2001 563 466.13 96.87 499.04 63.96 515.21 47.79

2002 584 495.19 88.80 537.41 46.59 558.22 25.77

2003 521.83 565.36 581.43

372.77 divided by 5 equals 74.55 258.85 divided by 5 equals 57.79
190.4690/5= 38.09

4.18Return resolved problem 4.1 and to the problems of 4.15 and 4.16 with
MAD as a criterion? I would use exponential smoothing with a constant of
smoothing of .3 as shown in problem 4.1 a moving average of 3
Would you use the trend to predict Volkswagen Beetle sales?
Explain your answer?
Soothing
Year Sales Exponential .3 MAD 3-Year Moving Average MAD
1998 450 410 40
1999 495 422 73
2000 518 444 74
2001 563 466 97 488 75
2002 584 495 89 525 59

MAD.. 373 divided by 5 equals 74.6 74.6 134/2= 67

The 3-year Moving Average MAD is 67.

MAD Exponential smoothing with a constant of .3 is 74.6
MAD Trend is 5.6 therefore it is the best

Year X Sales X2 XY Trend

1998 1 450 1 450 4.8
1999 2 495 4 990 6.6
2000 3 518 9 1,554 4
2001 4 563 16 2,252 7.4
2002 5 584 25 2,920 5.2
EX= 15 Ey=2,610 55 8,166 28
5 / 5 = 5.6

X = 15/5 = 3
 Y=2,610/5= 522

b= 8,166- (5) (3) (522) 336 33.6

55 - (5) (3"2) 10

a=522 - 33.6 (3) 421.2

The least squares equation for the trend is Y = 421.2 + 33.6x

4.19 the income in the law firm of Wesson for the period of
February to July have been?

F+ T+ FIT+
Income Trend Forecast
Me Annual Forecast per month smoothed including
Trend

Feb 70 65 0 65
Mar 68.5 70 + 1(70 - 65) = 70.5 1.9 72.4
Apr 64.8 70.5 + 0.1(68.5 - 70.8) = 70.3 1.56 71.86
May 71.7 70.3 + 0.1(64.8 - 70.3) = 69.75 1.36 71.11
June 71.3 69.75 + 0.1(71.7 - 69.75) = 69.95 1.13 71.08
Jul 72.8 69.95 + 0.1(71.3 - 69.95) = 70.09 0.93 71.02
Ago 70.09 + 0.1(72.8 - 70.09) = 70.36 0.8 69.56

March... = .2(70.5 - 65) + (1 - .2)(0)

= .2(5.5)+0.8
= 1.9

April... = .2(70.3-70.5)+(1-.2)(1.9)
= .2(0.2)+1.52
1.56

May... = .2(69.75-70.3)+(1-.2) (1.56)

= .2(0.55)+1.25
= 1.36
June... = 2(69.95 - 69.75) + (1 - 2)(1.36)
= .2(0.2)+1.088
= 1.13

Julio... = .2(70.09-69.95)+(1-.2)(1.13)
= .2(0.14)+0.904
= 0.93

August... = .2(70.36-70.09)+(1-.2)(0.93)
= .2(0.27)+0.744
0.80

4.20 solve problem 4.19 with a = .1 b = .8 using MSE as the constant

Does smoothing provide a better forecast?

4.21 see the illustration exponential smoothing with trend adjustment

from example 7 on pages 116 and 117 with a=0.2 and B=04, they forecast the
9 months of sales and we show the details in the calculations for months 2 and
3. With problem 4.2 we will continue the process for month 4.

In this problem, show your calculations for months 5 and 6 for Fr Tr and Fiar.

CALCULATIONS FOR MONTHS 5 AND 6

Step 1: F4 = .2(20) + (1 - .2)(15.18 + 2.10)

= 4 + (0.8)(17-28)

= 4 + 13.824= 17.82

Step 2: T4 = .4(17.82-15.18) + (1-.4)(2.10

= 0.4(2.64) + 0.6(2 - 10)

0.1 056 + 1.26 = 2.32

Step 3 FIT4= 17.82+2.32 =20.14

Step 1: F5= .4 (19.91-17.82) + (1-4) (2-32)
3.8 + (0.8) (20.14) = 19.91

Step 2: F5= .4(19.91-17.82) + (1-.4)(2-32)

. -15.252
0.84 + 1.39 = 2.23

Step 3: FIT5 = 19.91 + 2.23

22.14

4.22 return to problem 4.21 complete the forecast calculations of

exponential smoothing with trend adjustment for periods 7, 8, and 9
confirm that your figures for Fr Tr and FITr correspond to those in table 4.1

CALCULATION FOR MONTHS 7, 8, AND 9

Step 1: F7= .2(21)+(1-2) (22.51+2.38) Month 7

4.2 + 0.8(24.89)
= 24.11

Step 2: T7= .4(24.11-22.51) + (1-4) (2.38)

.4(1.6+) + .6(2.38)
0.64 + 1.43 = 2.07

Step 3: FIT7 = 24.11 + 2.07

=26.18

Step 1: F8= .2(31)+(1-.2)(24.11+2.07) My 8

6.2 + 0.8(26.18
27.14

Step 2: F8 = .4(27.14 - 24.11) + (1 - .4)(2.07)

2.22
1.21 + 1.24 = 2.45

Step 3: FIT = 27.14 + 2.45

29.59
 My 9
Step 1: F9 = .2(28) + (1-.2) (27.14+2.45)
5.6 + 0.8(29.59)
=29.28

Step 2: T9 = .4(29.28-27.14)+0.6(2.45)
.4 (2.14)+0.6(2.45)
0.85 + 1.47 = 2.32

Step 3: FIT9 = 29.28 + 2.32

=31.60

The table shows the sales of vegetable dehydrators during the

last year at the discount department store Bud Banis in St.
Louis. The managers prepared a forecast using a
combination of exponential smoothing and its collective judgment for the
following 4 months (March, April, May, and June of 2002)

Sales
Me Unitarians Forecasts MAD. MAPE
2001-2002 Adm.
July 100
August 93
September 96
October 110
November 124
December 119
January 92
February 83
March 101 120 19 100(19/101)=18.81 %
April 96 114 18 100(18/96)=18.75%
May 89 110 21 100(21/89)=23.60%
June 108 108 0 100(0/108)=0%

58 61.16%

MAD 58/4 = 14.5

 61.16%
MAPE 15.29%
4

The results of the Administration exceeded, they have MAD of 14.5 less than
MAPE
with 15.29 %

I would recommend MAD because its error is smaller 14.5

4.24

(Y) (x)
Appearance
Demand for pumps on TV. X2 Xy
of Green
Shades

3 3 9 9
6 4 16 24
7 7 49 49
5 6 36 30
10 8 64 80
8 5 25 40
1
EX=39 EY=33 99 32

Least squares regression to obtain a forecasting equation

X = 33/6 = 5.5

Y = 39/6 = 6.5

b = 232 - (6) (5.5) (6.5) 17.5 1

199-(6) (5.5"2) 17.5

a = 6.5 - 1 (5.5) 1

The least squares equation for the trend is Y = 1 + 1 x

The sales estimate for drums if Green Shades had appeared

9 times on TV last month
Y= 1 + 1 (9)

Y=1+9

Y = 10

Series of 10 Bombs

4.25

Month My (X) X2 XY
No. Of
Accident
Y s
1 January 30 900 30
2 February 40 1600 40
3 March 60 3600 60
4 April 90 8100 90
1
EY=10 220 4,200 20

Least squares regression to obtain

a forecasting equation

X = 220 / 4 = 55

Y = 10/4 = 2.5

b= 220- (4) (55)

(2.5) 330 0.16
14,200-(4)
55 inches 2 2100

a = 2.34
(55) 6.3

The least squares equation for the

the trend is Y = 2.5 + 6.3 x
4.26

Dema Deman
here you go Index
Prome Prome Station
Years god god nal
Station 2 Mensu
nes 1 2 Years to
Autumn 20 250 225 83 2.71
Summer 300 285 292.5 83 3.524
Winter 350 300 325 83 3.916
Primavera
ra 150 165 157.5 83 1.898

1,000.0
Total Average Annual Demand 0

Average Demand 1,000.00/

Monthly 12 months 83

Demand 1200
Next Year

Estacio
nes Demand
1200/12x2.
Autumn 710 = 271
1200/12x3.
Summer 524 = 352
Winter 1200 divided by 12 times 3.
 916= 392
Primave 1200/12x1.
ra 898 = 190

4.27
Days/Week Average Demand Average Daily Demand Seasonal Index
Monday 85 94 0.904
Tuesday 74 94 0.787
Wednesday 87 94 0.926
Thursday 97 94 1.032
Friday 134 94 1.423
Saturday 139 94 1.479
Sunday 42 94 -447

Total Average Demand

Weekly…………………………………………..658

Demand 658/7 days of the

Average week 94 Diaries

Diary

4.28
Average Demand
Seasons Quarters Average Demand Quarterly Index
2001 2002 2003
Winter 73 65 89 76 110 0.691
Spring 104 82 146 111 110 1.009
Summer 168 124 205 166 110 1.509
Autumn 74 52 222 87 110 0.791
Total average of
Demand 440/4 =
Quarterly 110

4.29
DEMAND CURVE D = 77 + 0.43Q

Quarter Index Factor

Winter 0.8
Spring 1.1
Summer 1.4
Autumn 0.7

FIRST QUARTER SECOND QUARTER THIRD QUARTER IV QUARTER

D = 77 + 0.43 (1) 77 + 0.43 (2) D = 77 + 0.43 (3) D = 77 + 0.43 (4)

D = 77 + 0.43 D= 77 + 0.86 D = 77 + 1.29 D = 77 + 1.72

D= 77.43 D= 77.86 D= 78.29 D= 78.72

Forecast for the four Quarters of 2005

Trimester Demand Kwts.

FIRST TRIMESTER 77 + 0.43(29) = 89.47
SECOND QUARTER 77 + 0.43(30) = 89.9
THIRD QUARTER 77 + 0.43(31) = 90.33
IV QUARTER 77 + 0.43(32) = 90.76

Adjusted Forecast
Quarter Demand kWts.
I TRIMESTER 71.58
 SECOND QUARTER (.8) (89.9) = 98.89
THIRD QUARTER 0.8 * 90.33 = 126.46
IV QUARTER 63.53

4.30

Y = 36 + 4.3x

Y= 36 + 4.3 (70)

Y= 337 The Aztec Demand

Temperature Demand
80°F

Y = 36 + 4.3 (80)

Y= 380

Temperature Demand
 90°F

Y = 36 + 4.3 (90)

Y= 423

4.31

(X) (Y)
Year Seasons Sales X2 Xy
1 Spring/Summer 26,825 1 26,825
2 Autumn/Winter 5,722 4 11,444
3 Spring/Summer 28,630 9 85,890
4 Autumn/Winter 7,633 16 30,532
5 Spring/Summer 30,255 25 151,272
6 Autumn/Winter 8,745 36 52,470
EX=21 107,810 91

Least Squares Regression to obtain a forecasting equation

X = 21/6 = 3.5

Y = 17,968.33 17,968.33

b=358,433- (6) (3.5) (17,968.33) 18,901.93 10,801.11

91-(6) (3.5"2) 17.5

a = 17,968.33 - 10,801.11 (3.5) 19,835.55

The least squares equation for the trend is Y = 10,801.11 + 19,835.55 x

Error

Sy,x = 2,622,101,548 - 19,835.55(107,810) - 10,801.11(358,433)

6 minus 2

Sy,x = 2,622,101,548 - 2,138,470,646 - 3,871,474.26

4

Sy,x = 479,759,427.7
4

Sy,x = 119,939,856.9

Sy,x = $ 10,951.71 Standard Error of Sales Estimation

4.32

(X) (Y)
Weeks Guests Sales of the Bar X2 Xy
1 16 $ 330 256 5,280
2 12 $270 144 3,240
3 18 $380 324 6,840
4 14 $300 196 4,200
n=4 60 1,280 920

Least squares regression to obtain a forecasting equation

X = 60 / 4 = 15
Y = 1,280 / 4 = 320

b=19,560- (4) (15) (320) 360.00 18

920-(4) (15"2) 20

a = 320 - 18 (15) 50.00

The least squares equation for the trend is Y = 50 + 18 x

Forecast is 20

Y = 20 + 18 (20) Y= 410

It is expected that
the sales are $410

4.33

(x) (y) errors

Year Transistors X2 XY Quadratic forecasts

1 140 1 1 80.36 59.64=3556.93

2 160 4 8 130.18 29.82=889.23

3 190 9 27 180 10=100

4 200 16 64 229.82 (29.82)=889.23

5 210 25 125 279.64 (-69.64)=4879.73

15 EY = 900 55 10,285.12

MAPE

100(59.64/140)=42.6

18.64

100(10/190)=5.26
14.91

100(69.64/210)=33.16

114.57%

X = 15/5
=3

Y= 900/5= 180

b= 225- (3) (3) (180) 1,395.00 49.82

55 - (3) (3'2) 28

a = 180 - 49.82 (3) 30.54

The least squares equation for the trend is Y = 30.54 +

49.82 x
Transfers for the Next
year

Y = 30.54 + 49.82
Y= 30.54 + 49.82 (1) Y= 30.54 + 49.82 (2) Y= 30.54 + 49.82 (3) (4)
Y=
Y = 80.36 130.18 Y=180 Y= 229.82

Mean Squared Error MSE using Regression

Linear

2,057.02
NEXT YEAR
Y= 30.54 + 49.82
(6)
Mean Absolute Percentage Error (MAPE) Y= 329.46

114.57% 22.91%
5
4.34
No, of Amount of
Automobiles Beverage Sales rain
Registered in Alcoholic beverages
Miles units inches
X1 X2 X3
(a) 2 3 0
(b) 3 5 1
(c) 4 7 2

FORMULA

Y = a + b1 x1 + b2 x2 + b3 x3

Where Y = Number of Automobile Accidents

a=
7.5

b1=
3.5

b2=
4.5

b3=
2.5

CALCULATE THE NUMBER OF EXPECTED ACCIDENTS IN THE

CONDITIONS A, B, C

A)
Y = 7.5 + 3.5 (2) + 4.5 (3) + 2.5
0

Y= 7.5 + 7 + 13.5 + 0

Y= 28

B)
Y= 7.5 + 3.5 (3) + 4.5 (5) + 2.5
(1)

Y = 7.5 + 10.5 + 22.5 + 2.5

Y= 43

C)
Y = 7.5 + 3.5 (4) + 4.5 (7) + 2.5
(2)

Y = 7.5 + 14 + 31.5 + 5

Y= 58