=========================preview======================
(ISMT162)2006fall_Quiz_1_ISMT162_A[1].pdf
Back to ISMT162 Login to download
======================================================
Quiz 1
Time: 9:00am C 9:50am
(9 Questions, 20 Points)

Note: Each question has 5 choices. There is only ONE correct answer for each question. You are required to write down the reason or the procedure to find your choice on the sheet. If you only give the correct answer without reasonable explanation or the procedure, you obtain 0.5 point for each question.

Name: Student No.:


Question 1. (3 points) An electrical contractors records during the last four weeks indicate the number of job requests:
Week
1
2
3
4

Request
17
19
16
20



The contractor first chooses a forecast method. Then, the contractor starts predicting the number of requests for week 5. It turns out that the number of requests in week 5 is 18. After knowing it, the contract continues doing forecast for week 6 by the same method.
Finally, the number of requests in week 6 is realized as 22. The previous forecast for week 4 is 22.

Which method is the most accurate by the measures of MSE and MAD for the forecast errors in week 5 and 6?
(a) Na.ve;
(b) A 3-period moving average;
(c) A 4-period moving average;
(d) Exponential smoothing with =0.1;
(e) Exponential smoothing with =0.3.

Answer: c
Reason or Procedure:



(a), The forecasts for period 5 and 6 are 20 and 18, respectively.
The related forecast errors are 18 - 20 = -2 and 22 -18 = 4

(b) The related forecast errors are 18-(19+16+20)/3=0.3333 and
22- (16+20+18)/3=4, respectively.

(c) The related forecast errors are 18 -(17+19+16+20)/4=0 and
22-(19+16+20+18)/4=3.75, respectively.

(d) The related forecast errors are 18 C (20*0.1+0.9*22)= 18-21.8=-3.8 and
22- (18*0.1+0.9*21.8)=22-21.42=0.58, respectively.


(e) The related forecast errors are 18- (20*0.3+0.7*22)=21.4 and
22- (18*0.3+0.7*21.4)=22-20.38 =1.62, respectively.

Method
MAD
MSE

(a)
3
20

(b)
2.1667
16.11111

(c)
1.875
14.0625

(d)
2.19
14.7764

(e)
2.51
14.1844




From the computational results in the table, (c) is the best.





Question 2. (1 Point) The cumulative forecast error is important for determining the
(a) Mean squared error;
(b) Bias in forecast error;
(c) The tend equation;
(d) Mean absolute percent deviation;
(e) Control limits;

Answer: b
Reason or Procedure:


Because the cumulative forecast error is given by
, it represents the bias in forecast error.












Question 3. (3 points) The following diagram describes a process that consists of eight separate operations, with sequential relationships and capacities (units per hour) as show. If you could increase the capacity of only two operations through process i