Scribd
Upload
53 views5 pages

Time Series Forecasting Exercise: Simple Exponential Smoothing

This document describes using simple exponential smoothing to forecast time series sales data. It shows the sales data over 10 periods and calculates the forecast using simple exponential smoothing with different alpha values. It uses the Excel Solver tool to find the optimal alpha value that minimizes the mean absolute deviation (MAD) error between actual and forecast sales. The optimal alpha is found to be 0.80094331, resulting in a minimum MAD of 9.08.

Uploaded by

Pravin K Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLS, PDF, TXT or read online on Scribd
53 views5 pages

Time Series Forecasting Exercise: Simple Exponential Smoothing

This document describes using simple exponential smoothing to forecast time series sales data. It shows the sales data over 10 periods and calculates the forecast using simple exponential smoothing with different alpha values. It uses the Excel Solver tool to find the optimal alpha value that minimizes the mean absolute deviation (MAD) error between actual and forecast sales. The optimal alpha is found to be 0.80094331, resulting in a minimum MAD of 9.08.

Uploaded by

Pravin K Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLS, PDF, TXT or read online on Scribd
You are on page 1/ 5

Time Series Forecasting Exercise

Simple Exponential Smoothing


alpha=

Period
1
2
3
4
5
6
7
8
9
10

Sales
99
88
50
58
62
60
55
62
71
65

Simple
Exponential
Smoothing
N/A
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0

Error Abs. Error


-11.0
-49.0
-41.0
-37.0
-39.0
-44.0
-37.0
-28.0
-34.0
-35.56
BIAS

11.0
49.0
41.0
37.0
39.0
44.0
37.0
28.0
34.0
35.56
MAD

Objective is to minimize
the error. In this case
MAD
Standard Error (Square Root of MSE) =

Select a value of alpha


that will minimize MAD

Percent
Error

Squared
Error

12.50%
98.00%
70.69%
59.68%
65.00%
80.00%
59.68%
39.44%
52.31%
59.70%
MAPE

121.0
2401.0
1681.0
1369.0
1521.0
1936.0
1369.0
784.0
1156.0
1370.89
MSE
37.0

Time Series Forecasting Exercise

Select Tools - Solver

Simple Exponential Smoothing


alpha=

Period
1
2
3
4
5
6
7
8
9
10

Sales
99
88
50
58
62
60
55
62
71
65

Simple
Exponential
Smoothing
N/A
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0

Error Abs. Error


-11.0
-49.0
-41.0
-37.0
-39.0
-44.0
-37.0
-28.0
-34.0
-35.56
BIAS

11.0
49.0
41.0
37.0
39.0
44.0
37.0
28.0
34.0
35.56
MAD

Objective is to minimize
the error. In this case
MAD
Standard Error (Square Root of MSE) =

Percent
Error

Squared
Error

12.50%
98.00%
70.69%
59.68%
65.00%
80.00%
59.68%
39.44%
52.31%
59.70%
MAPE

121.0
2401.0
1681.0
1369.0
1521.0
1936.0
1369.0
784.0
1156.0
1370.89
MSE
37.0

Set the target cell equal to the


value of MAD (cell E19) and sele
minimize. Set the By Changing
Cells equal to alpha (cell E4)

Set the target cell equal to the


value of MAD (cell E19) and select
minimize. Set the By Changing
Cells equal to alpha (cell E4)

Time Series Forecasting Exercise


Simple Exponential Smoothing

Optimal value of alpha

alpha= 0.80094331

Period
1
2
3
4
5
6
7
8
9
10

Sales
99
88
50
58
62
60
55
62
71
65

Simple
Exponential
Smoothing
N/A
99.0
90.2
58.0
58.0
61.2
60.2
56.0
60.8
69.0

Minimum value of MAD

Error Abs. Error

Percent
Error

Squared
Error

-11.0
-40.2
0.0
4.0
-1.2
-5.2
6.0
10.2
-4.0
-4.61
BIAS

12.50%
80.38%
0.00%
6.45%
2.01%
9.53%
9.61%
14.35%
6.11%
15.66%
MAPE

121.0
1615.2
0.0
16.0
1.4
27.5
35.5
103.8
15.8
215.13
MSE

11.0
40.2
0.0
4.0
1.2
5.2
6.0
10.2
4.0
9.08
MAD

Standard Error (Square Root of MSE) =

14.7

After Solver finds a solution


select OK to keep the solution.
Note the value in E4 = 0.8009433
and the MAD = 9.08. This is the
minimum error for this time serie
forecast.

After Solver finds a solution


select OK to keep the solution.
Note the value in E4 = 0.8009433
and the MAD = 9.08. This is the
minimum error for this time series
forecast.

You might also like