Scribd
Upload
75 views

SCM Question Set+ Solutions

The document contains 5 questions regarding forecasting techniques such as moving averages, exponential smoothing, regression analysis, economic order quantity model, and analyzing forecast accuracy. Question 1 computes 3-month and 6-month moving average forecasts and notes that increasing the time horizon from 3 to 6 months results in smoother forecasts. Question 2 initializes an exponential smoothing forecast using the average of historical data and compares the forecasts to a 6-month moving average. Question 3 compares the accuracy of two forecasting methods using MAD, MSE, and MAPE and finds they do not always indicate the same best method. Question 4 determines the economic order quantity and reorder point for imported car mufflers, noting the current policy incurs additional costs. Question 5 uses regression analysis

Uploaded by

Xiu
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
75 views

SCM Question Set+ Solutions

The document contains 5 questions regarding forecasting techniques such as moving averages, exponential smoothing, regression analysis, economic order quantity model, and analyzing forecast accuracy. Question 1 computes 3-month and 6-month moving average forecasts and notes that increasing the time horizon from 3 to 6 months results in smoother forecasts. Question 2 initializes an exponential smoothing forecast using the average of historical data and compares the forecasts to a 6-month moving average. Question 3 compares the accuracy of two forecasting methods using MAD, MSE, and MAPE and finds they do not always indicate the same best method. Question 4 determines the economic order quantity and reorder point for imported car mufflers, noting the current policy incurs additional costs. Question 5 uses regression analysis

Uploaded by

Xiu
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

SCM Question Set

1. Observations of the demand for a certain part stocked at a warehouse during year 1985
were:

Month Deman Month Deman


d d
January 89 July 223
Februar 57 August 286
y
March 144 Septembe 212
r
April 221 October 275
May 177 Novembe 188
r
June 280 Decembe 312
r

Compute the one-step-ahead three-month and six-month moving-average forecasts for July
through December. What effect does increasing N from 3 to 6 have on the forecasts?

2. For the data in Q1, use the arithmetic average of the first six months of data as a baseline
to initialize the exponential smoothing.
a) Determine the one-step-ahead exponential smoothing forecasts for August through
December, assuming α = 0.20.
b) Compare the accuracy of the forecasts obtained in part (a) with the one-step-ahead six-
month moving-average forecasts

3. Two forecasting methods have been used to evaluate the same economic time series. The
results are:
Forecast from Method 1 Forecast from Method 2 Realized Value of the Series
223 210 256
289 320 340
430 390 375
134 112 110
190 150 225
550 490 525
Compare the effectiveness of these methods by computing the MSE, MAD, and the MAPE.
Does each of the measures of forecasting accuracy indicate that the same forecasting
technique is best? If not, why?

4. A large automobile repair shop installs about 1250 mufflers (a device that absorbs noise,
especially one used with an internal-combustion engine) per year, 18 percent of which are for
imported cars. All of the imported-car mufflers are purchased from a single local supplier at a
cost of $18.50 each. The shop uses a holding cost based on a 25 percent annual interest rate.
The setup cost for placing an order is estimated to be $28.
a) Determine the optimal number of imported-car mufflers the shop should purchase each
time an order is placed, and the time between placement of orders.
b) If the replenishment lead time is six weeks, what is the reorder point based on the level of
on-hand inventory?
c) The current reorder policy is to buy imported-car mufflers only once a year. What is the
additional holding and setup cost incurred by this policy?

5. A national park, has kept close tabs on the number of guide tours using the park since its
opening in January 1993. For the first six months of operation, the following figures were
recorded:

Month # of Tours Month # of Tours


January 133 April 640
February 183 May 1876
March 285 June 2550

a) Draw a graph of these six data points.


Assume that January = period 1, February = period 2, and so on.. Estimate the slope and
intercept from your graph.
b) Compute the exact values of the intercept a and the slope b from the regression equations.
c) What are the forecasts obtained for July through December 1993 from the regression
equation determined in part (b)?
d) Comment on the results you obtained in part (c). Specifically, how confident would you be
about the accuracy of the forecasts that you obtained?
Q1)
Month Deman MA(3 MA(6
d ) )
January 89
February 57
March 144
April 221
May 177
June 280
July 223 226 161.33
August 286 226.67 183.67
Septemb 212 263 221.83
er
October 275 240.33 233.17
Novembe 188 257.67 242.17
r
Decembe 312 225 244
r

Q2) The arithmetic average of the first six months is 968 / 6 = 161.33. We will use this as a
baseline to initialize the exponential smoothing. We assume that 161.33 is the forecast for
July.
(a & b )
New forecast = α (Current observation of demand) + (1 – α) (Last forecast)
or
Ft =α Dt−1 +(1−α ) Ft−1
⇒ F 2 =(0 . 2) (223 )+(0 . 8) (161 .33 )=173 . 664
⇒ F 3 =(0 . 2) (286 )+(0 . 8) (173 .664 )=196 . 131
And so on;
Deman
Month ES(0.2) Error MA(6) Error
d
August 286 173.664 -112.34 183.67 -102.33
September 212 196.131 -15.87 221.83 9.83
October 275 199.305 -75.70 233.17 -41.83
November 188 214.444 26.44 242.17 54.17
December 312 209.155 -102.84 244 -68.00

For exponential smoothing;


n 5
1 1 333 . 19
MAD= ∑ |e i|= ∑ |e i|= =66 .634
n i=1 5 i=1 5
For moving average MA(6);
n 5
1 1 276 . 167
MAD= ∑ |e i|= ∑ |e i|= =55 .233
n i=1 5 i=1 5
According to the MAD values, MA(6) gave more accurate forecasts.
Q3)
Forecast 1 Demand Error 1 Abs(Error 1) sqr(Error 1) Error1/Demand
223 256 -33 33 1089 0,1289
289 340 -51 51 2601 0,1500
430 375 55 55 3025 0,1467
134 110 24 24 576 0,2182
190 225 -35 35 1225 0,1556
550 525 25 25 625 0,0476

n 6
1 1 223
MAD 1 = ∑ |ei|= ∑ |ei|= =37 .167
n i =1 6 i =1 6
n 6
1 1 9141
MSE1 = ∑ e2i = ∑ e2i = =1523 .5
n i=1 6 i=1 6

[ ] [ ]
n 6
1 ei 1 ei 0. 8469
MAPE1 = ∑ | | ×100= ∑ | | ×100= ×100=14 . 115
n i=1 Di 6 i=1 D i 6

Demand Error 2 Abs(Error 2) sqr(Error 2) Error2/Demand


Forecast 2
210 256 -46 46 2116 0,1797
320 340 -20 20 400 0,0588
390 375 15 15 225 0,0400
112 110 2 2 4 0,0182
150 225 -75 75 5625 0,3333
490 525 -35 35 1225 0,0667

n 6
1 1 193
MAD2 = ∑ |e i|= ∑ |e i|= =32. 167
n i=1 6 i =1 6
n 6
1 1 9595
MSE 2 = ∑ e 2i = ∑ e 2i = =1599. 167
n i=1 6 i=1 6

[ ] [ ]
n 6
1 ei 1 e 0 . 6966928
MAPE 2 = ∑ | | ×100= ∑ | i | ×100= ×100=11 . 612
n i=1 Di 6 i=1 Di 6

The measures of MAD, MSE, and MAPE does not show that the same forecasting technique
is best. The reason for this may be that the sensitiveness of the measures for the forecast
accuracy techniques are not same for errors.
Q4) D = (1250) (0.18) = 225
c = $18.50; I = 0.25; A = $28; h = Ic = (0.25) ($18.50) = $4.625

a)
EOQ=Q ¿=

¿
√ 2 AD
h
=
√2(28 )(225)
( 4 . 625)
=52

Q 52
T= = =0 .2311 years
D 225 =12.0172 weeks

b) Reorder point = (lead time) (demand) = (6 weeks) (225 / 52) = 25.96  26 units

c) For the current reorder policy (buy only once a year)


TC = annual setup cost + annual holding cost
AD hQ ( $ 28 )(225 ) ($ 4 .625 )(225 )
+ = + =$ 28+$ 520 .31=$ 548. 31
= Q 2 225 2

For the optimal policy


TC = annual setup cost + annual holding cost
AD hQ ( $ 28 )(225 ) ($ 4 .625 )(52)
+ = + =$ 121. 15+$ 120 .25=$ 241. 40
= Q 2 52 2

As we see, the annual difference between the two policies is:


$548.31 - $241.40 = $306.91
Q5)
a) The graph for these six data points is;
3000

2500

2000
Demand

1500 Demand

1000

500

0
January February March April May June
Month

Estimates for the slope and the intercept are:


Slope = 2800 / 6 = 466.67; Intercept = -500
b) Sxy = (6) (28594) – (21) (5667) = 52557 and Sxx = (6) (91) – (21)2 =105
S xy 52557 D−b ( n+1) (944 . 5 )−(500 .54 )(7 )
b= = =500 .54 a= = =−807 . 4
S xx 105 and 2 2

c)
D^ t =−807 .4+(500 .54) t
t Month Forecas
t
7 July 2696.38
8 August 3196.92
9 Septembe 3697.46
r
10 October 4198
11 Novembe 4698.54
r
12 Decembe 5199.08
r

d) There may be peak usage in some months. Moreover, the increasing trend will possibly not
continue indefinitely.

You might also like