Scribd
Upload
147 views11 pages

Double Exponential Smoothing Analysis

The document discusses using double exponential smoothing to forecast monthly sales data from 2018 and 2019. It tests different combinations of alpha and beta values (0.1, 0.2, 0.3) and measures the forecast accuracy using mean squared error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The results show that an alpha of 0.3 and beta of 0.3 generally provided the most accurate forecasts between the two time periods according to all three error metrics.

Uploaded by

Shahnewaz Parvez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
147 views11 pages

Double Exponential Smoothing Analysis

The document discusses using double exponential smoothing to forecast monthly sales data from 2018 and 2019. It tests different combinations of alpha and beta values (0.1, 0.2, 0.3) and measures the forecast accuracy using mean squared error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The results show that an alpha of 0.3 and beta of 0.3 generally provided the most accurate forecasts between the two time periods according to all three error metrics.

Uploaded by

Shahnewaz Parvez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

Double Exponential

Smoothing
Double Exponential Smoothing
Tt (initial) = ((Lt+1 – Lt) + (Lt+3 – Lt-2))/2
Tt = Beta * (At – A) + (1-Beta) * Tt-1
At = Alpha * Tt + (1-Alpha) * (At (initial) – Tt (initial))
Forecast = At – Tt
Analysis & Results of 2018
When Alpha=0.1, Beta=0.1
Monthly Sales of the year 2018
2500000

2000000

1500000

1000000

500000

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

Net Sales / Taka Forecast


Analysis & Results of 2018
When Alpha=0.2, Beta=0.2
Monthly Sales of the year 2018
2000000

1800000

1600000

1400000

1200000

1000000

800000

600000

400000

200000

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

Net Sales / Taka Forecast


Analysis & Results of 2018
When Alpha=0.3, Beta=0.3
Monthly Sales of the year 2018
1800000

1600000

1400000

1200000

1000000

800000

600000

400000

200000

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

Net Sales / Taka Forecast


Analysis & Results of 2019
When Alpha=0.1, Beta=0.1
Monthly Sales of the year 2019
1600000

1400000

1200000

1000000

800000

600000

400000

200000

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

Net Sales / Taka Forecast Series3


Analysis & Results of 2019
When Alpha=0.2, Beta=0.2
Monthly Sales of the year 2019
1600000

1400000

1200000

1000000

800000

600000

400000

200000

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

Net Sales / Taka Forecast


Analysis & Results of 2019
When Alpha=0.3, Beta=0.3
Monthly Sales of the year 2019
1600000

1400000

1200000

1000000

800000

600000

400000

200000

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

Net Sales / Taka Forecast


Measuring Forecast Error
• Measuring Forecast Error
Error
­ = Actual Data-Forecast
The Comparison of
MSE
For Data of 2018:
MSE 0.1 5.60654E+11

MSE 0.2 2.45345E+11

MSE0.3 1.05552E+11

For Data of 2019:

MSE 0.1 67403833765

MSE 0.2 71372406196

MSE0.3 56669980686

0.3 Is the better forecast


The Comparison of
MAE
For Data of 2018:
MAE 0.1 616819.2189

MAE 0.2 432990.5976

MAE0.3 293161.5835

For Data of 2019:

MAE 0.1 216953.3469

MAE 0.2 218568.7642

MAE0.3 212226.824
The Comparison of
MAPE
For Data of 2018:
MAPE 0.1 61.87%

MAPE 0.2 42.35%

MAPE0.3 27.76%

For Data of 2019:

MAPE 0.1 20.37%

MAPE 0.2 20.91%

MAPE0.3 19.79%

You might also like