FOREC

AST AC
CURAC
Measures used to evaluate
forecast accuracy
By: Sheila Mae Laceron

Y
a
Forecas 2

Where :

t e = forecast error
Xt = actual data at perio

Error dt
Ft = forecast at period t

e=Xt-Ft
 3

Measures Mean Absolute Deviation

(MAD)
Average of the difference between the forecast and the a

used to ctual data

Mean Absolute Percentage E
rror (MAPE)
evaluate
Measures the absolute error in terms of percentage rath
er than of period

Cumulative Error
forecast
Sum of the forecast error

Average Error
accuracy
Average of the cumulative error per period
MEAN
ABSOL Xt = data in period t

UTE
Ft = forecast in period t
n = total number of periods
| | = absolute value

DEVIAT
 Demand ES (Ft) Error ES (Ft) Error Error Error
Period |Xt-Ft| |Xt-Ft| AES |Xt-Ft| LTL |Xt-Ft|
Xt = 30 Xt-Ft = 50 Xt-Ft Xt-Ft Xt-Ft
1 47 -- -- -- -- -- -- -- --
46.96 -- -- --
2 50 47.00 3.00 3.00 47.00 3.00 3.00 47.00 3.00
48.68 3.0 1.32 1.32
3 51 47.90 3.10 3.10 48.50 2.50 2.50 48.95 2.05
50.40 2.05 0.60 0.60
4 47 48.83 -1.83 1.83 49.75 -2.75 2.75 50.44 -3.44
52.12 3.44 5.12 -5.12
5 55 48.28 6.72 6.72 48.38 6.62 6.62 48.44 6.56
53.84 6.56 1.16 1.16
6 60 50.30 9.70 9.70 51.69 8.31 8.31 52.73 7.27
55.56 7.27 4.44 4.44
7 53 53.21 -0.21 0.21 55.85 -2.85 2.85 57.82 -4.82
57.28 4.82 4.28 -4.28
8 57 53.15 3.85 3.85 54.43 2.57 2.57 55.37 1.63
59.00 1.63 2.00 -2.00
9 66 54.30 11.70 11.70 55.72 10.28 10.28 56.76 9.24
60.72 9.24 5.28 5.28
10 62 57.81 4.19 4.19 60.86 1.14 1.14 63.13 -1.13
62.44 1.13 0.44 -0.44
11 65 59.07 5.93 5.93 61.43 3.57 3.57 63.19 1.81
64.16 1.81 0.84 0.84
12 64 60.85 3.15 3.15 63.22 0.78 0.78 64.98 -0.98
65.88 0.98 1.88 -1.88
677.0
630 580.70 49.30 53.38 596.83 33.17 44.37 608.81 21.19 41.93 0.00 27.36
4

Computing for MAD :

Exponential smoothing ( = 30) : Adjusted Exponential Smoothing:

Exponential smoothing ( = 50) : Linear Trend Line :

 Demand ES (Ft) Error ES (Ft) Error Error Error
Period |Xt-Ft| |Xt-Ft| AES |Xt-Ft| LTL |Xt-Ft|
Xt = 30 Xt-Ft = 50 Xt-Ft Xt-Ft Xt-Ft
1 47 -- -- -- -- -- -- -- --
46.96 -- -- --
2 50 47.00 3.00 3.00 47.00 3.00 3.00 47.00 3.00
48.68 3.0 1.32 1.32
3 51 47.90 3.10 3.10 48.50 2.50 2.50 48.95 2.05
50.40 2.05 0.60 0.60
4 47 48.83 -1.83 1.83 49.75 -2.75 2.75 50.44 -3.44
52.12 3.44 5.12 -5.12
5 55 48.28 6.72 6.72 48.38 6.62 6.62 48.44 6.56
53.84 6.56 1.16 1.16
6 60 50.30 9.70 9.70 51.69 8.31 8.31 52.73 7.27
55.56 7.27 4.44 4.44
7 53 53.21 -0.21 0.21 55.85 -2.85 2.85 57.82 -4.82
57.28 4.82 4.28 -4.28
8 57 53.15 3.85 3.85 54.43 2.57 2.57 55.37 1.63
59.00 1.63 2.00 -2.00
9 66 54.30 11.70 11.70 55.72 10.28 10.28 56.76 9.24
60.72 9.24 5.28 5.28
10 62 57.81 4.19 4.19 60.86 1.14 1.14 63.13 -1.13
62.44 1.13 0.44 -0.44
11 65 59.07 5.93 5.93 61.43 3.57 3.57 63.19 1.81
64.16 1.81 0.84 0.84
12 64 60.85 3.15 3.15 63.22 0.78 0.78 64.98 -0.98
65.88 0.98 1.88 -1.88
677.0
630 580.70 49.30 53.38 596.83 33.17 44.37 608.81 21.19 41.93 0.00 27.36
4

Computing for MAPE :

Exponential smoothing ( = 30) : Adjusted Exponential Smoothing:

Exponential smoothing ( = 50) : Linear Trend Line :

 Demand ES (Ft) Error ES (Ft) Error Error Error
Period |Xt-Ft| |Xt-Ft| AES |Xt-Ft| LTL |Xt-Ft|
Xt = 30 Xt-Ft = 50 Xt-Ft Xt-Ft Xt-Ft
1 47 -- -- -- -- -- -- -- --
46.96 -- -- --
2 50 47.00 3.00 3.00 47.00 3.00 3.00 47.00 3.00
48.68 3.0 1.32 1.32
3 51 47.90 3.10 3.10 48.50 2.50 2.50 48.95 2.05
50.40 2.05 0.60 0.60
4 47 48.83 -1.83 1.83 49.75 -2.75 2.75 50.44 -3.44
52.12 3.44 5.12 -5.12
5 55 48.28 6.72 6.72 48.38 6.62 6.62 48.44 6.56
53.84 6.56 1.16 1.16
6 60 50.30 9.70 9.70 51.69 8.31 8.31 52.73 7.27
55.56 7.27 4.44 4.44
7 53 53.21 -0.21 0.21 55.85 -2.85 2.85 57.82 -4.82
57.28 4.82 4.28 -4.28
8 57 53.15 3.85 3.85 54.43 2.57 2.57 55.37 1.63
59.00 1.63 2.00 -2.00
9 66 54.30 11.70 11.70 55.72 10.28 10.28 56.76 9.24
60.72 9.24 5.28 5.28
10 62 57.81 4.19 4.19 60.86 1.14 1.14 63.13 -1.13
62.44 1.13 0.44 -0.44
11 65 59.07 5.93 5.93 61.43 3.57 3.57 63.19 1.81
64.16 1.81 0.84 0.84
12 64 60.85 3.15 3.15 63.22 0.78 0.78 64.98 -0.98
65.88 0.98 1.88 -1.88
677.0
630 580.70 49.30 53.38 596.83 33.17 44.37 608.81 21.19 41.93 0.00 27.36
4

Computing for Cumulative Error : where

Exponential smoothing ( = 30) : Adjusted Exponential Smoothing:

Exponential smoothing ( = 50) :

 Demand ES (Ft) Error ES (Ft) Error Error Error
Period |Xt-Ft| |Xt-Ft| AES |Xt-Ft| LTL |Xt-Ft|
Xt = 30 Xt-Ft = 50 Xt-Ft Xt-Ft Xt-Ft
1 47 -- -- -- -- -- -- -- --
46.96 -- -- --
2 50 47.00 3.00 3.00 47.00 3.00 3.00 47.00 3.00
48.68 3.0 1.32 1.32
3 51 47.90 3.10 3.10 48.50 2.50 2.50 48.95 2.05
50.40 2.05 0.60 0.60
4 47 48.83 -1.83 1.83 49.75 -2.75 2.75 50.44 -3.44
52.12 3.44 5.12 -5.12
5 55 48.28 6.72 6.72 48.38 6.62 6.62 48.44 6.56
53.84 6.56 1.16 1.16
6 60 50.30 9.70 9.70 51.69 8.31 8.31 52.73 7.27
55.56 7.27 4.44 4.44
7 53 53.21 -0.21 0.21 55.85 -2.85 2.85 57.82 -4.82
57.28 4.82 4.28 -4.28
8 57 53.15 3.85 3.85 54.43 2.57 2.57 55.37 1.63
59.00 1.63 2.00 -2.00
9 66 54.30 11.70 11.70 55.72 10.28 10.28 56.76 9.24
60.72 9.24 5.28 5.28
10 62 57.81 4.19 4.19 60.86 1.14 1.14 63.13 -1.13
62.44 1.13 0.44 -0.44
11 65 59.07 5.93 5.93 61.43 3.57 3.57 63.19 1.81
64.16 1.81 0.84 0.84
12 64 60.85 3.15 3.15 63.22 0.78 0.78 64.98 -0.98
65.88 0.98 1.88 -1.88
677.0
630 580.70 49.30 53.38 596.83 33.17 44.37 608.81 21.19 41.93 0.00 27.36
4

where , n = number of period

Computing for Average Error :
Exponential smoothing ( = 30) : Adjusted Exponential Smoothing:

Exponential smoothing ( = 50) :

JUDGMENTAL FOREC
9
ASTING
Uses managerial judgment, expertise and opinions

Judgmental forecasting
educated guess by the sales managers to a consen
sus forecast of the senior executives of the company

Sales force estimates

The sum of individual sales peoples opinion based on his sales e
xperience and on his estimate of sale in the future

Customers expectations
Based on surveys, formal or informal of planned purchas
es