Chinanart
 
เลือกไซต์นี้

การนำทาง
Home‎ > ‎

Project_1

For project 1, I picked the datasets from the cases in the book. There are 3 related datasets in the file.

 

I started by forecasting a labor hour of the first week of 2003 using 4 period moving average using week period. As a result, StatTools calculated a mean absolute error of 223.27 and root mean square error of 270.94, and a percentage absolute error of 41.43%, which, in my opinion is fairly high.

 

 

Moving Averages

 

Mean Abs Err

223.27

Root Mean Sq Err

270.94

Mean Abs Per% Err

41.63%

 

StateTools also created a graph to present the differences between actual data and forecasting data.

 

 

The forecasting data imply that labor hours of the company will be lower. However, after adding a power trend line on forecasting data (a yellow line), it shows that labor hours are lower at the beginning of 2002. Then labor hours tend to increase and become steady toward the end of the year.

 

 

 

 

I also create a power trend line in an actual data to see the differences between these two trend lines. As a result, there is not much difference between the two trend lines.

 

Forecasting Data

LaborHrs

Forecast

Error

Week1-2002

695.0000

Week2-2002

531.0000

Week3-2002

839.0000

Week4-2002

379.0000

Week5-2002

602.0000

611.00

-9.00

Week6-2002

807.0000

587.75

219.25

Week7-2002

515.0000

656.75

-141.75

Week8-2002

851.0000

575.75

275.25

Week9-2002

774.0000

693.75

80.25

Week10-2002

766.0000

736.75

29.25

Week11-2002

537.0000

726.50

-189.50

Week12-2002

642.0000

732.00

-90.00

Week13-2002

434.0000

679.75

-245.75

Week14-2002

741.0000

594.75

146.25

Week15-2002

995.0000

588.50

406.50

Week16-2002

675.0000

703.00

-28.00

Week17-2002

935.0000

711.25

223.75

Week18-2002

638.0000

836.50

-198.50

Week19-2002

1128.0000

810.75

317.25

Week20-2002

744.0000

844.00

-100.00

Week21-2002

180.0000

861.25

-681.25

Week22-2002

188.0000

672.50

-484.50

Week23-2002

755.0000

560.00

195.00

Week24-2002

1014.0000

466.75

547.25

Week25-2002

371.0000

534.25

-163.25

Week26-2002

758.0000

582.00

176.00

Week27-2002

662.0000

724.50

-62.50

Week28-2002

721.0000

701.25

19.75

Week29-2002

811.0000

628.00

183.00

Week30-2002

944.0000

738.00

206.00

Week31-2002

1184.0000

784.50

399.50

Week32-2002

1024.0000

915.00

109.00

Week33-2002

772.0000

990.75

-218.75

Week34-2002

635.0000

981.00

-346.00

Week35-2002

630.0000

903.75

-273.75

Week36-2002

417.0000

765.25

-348.25

Week37-2002

665.0000

613.50

51.50

Week38-2002

682.0000

586.75

95.25

Week39-2002

1046.0000

598.50

447.50

Week40-2002

887.0000

702.50

184.50

Week41-2002

706.0000

820.00

-114.00

Week42-2002

520.0000

830.25

-310.25

Week43-2002

813.0000

789.75

23.25

Week44-2002

387.0000

731.50

-344.50

Week45-2002

444.0000

606.50

-162.50

Week46-2002

725.0000

541.00

184.00

Week47-2002

875.0000

592.25

282.75

Week48-2002

1026.0000

607.75

418.25

Week49-2002

407.0000

767.50

-360.50

Week50-2002

815.0000

758.25

56.75

Week51-2002

706.0000

780.75

-74.75

Week52-2002

1231.0000

738.50

492.50

Week1-2003

789.75

 

The forecast of labor hours in the first week of 2003 is 789.75 which is about 94.75 hours different with 41.63 percent error.

 

ไฟล์แนบ (1)