Week Value 3-Week Moving Average Forecast Error Absolute Forecast Error

1 240
2 350
3 230
4 260 273 -13.33 13.33
5 280 280 0.00 0.00
6 320 257 63.33 63.33
7 220 287 -66.67 66.67
8 310 273 36.67 36.67
9 240 283 -43.33 43.33
10 310 257 53.33 53.33
11 240 287 -46.67 46.67
12 230 263 -33.33 33.33

𝑦 ̂_(𝑡+1)=(𝑦_𝑡+𝑦MFE
_(𝑡−
13 260

1)+𝑦_(𝑡−2))/3
MAE

-5.56 39.63

w1 0.5
w2 0.3
w3 0.2
Week Value Weighted Moving Average Forecast Error Absolute Forecast Error
1 240
2 350
3 230
4 260 268 -8.00 8.00
5 280 269 11.00 11.00
6 320 264 56.00 56.00
7 220 296 -76.00 76.00
8 310 262 48.00 48.00
9 240 285 -45.00 45.00
10 310 257 53.00 53.00
11 240 289 -49.00 49.00
12 230 261 -31.00 31.00
13 249

𝑦 ̂_(𝑡+1)=𝑤_1 𝑦_𝑡+𝑤_2
𝑦_(𝑡−1)+𝑤_3 𝑦_(𝑡−2)
MFE MAE

-4.56 41.89

α 0.2
Week Value Exponential Smoothing Forecast Error Absolute Forecast Error
1 240 240 0.0000 0.0000
2 350 240 110.0000 110.0000
3 230 262 -32.0000 32.0000
4 260 256 4.4000 4.4000
5 280 256 23.5200 23.5200
6 320 261 58.8160 58.8160
7 220 273 -52.9472 52.9472
8 310 262 47.6422 47.6422
9 240 272 -31.8862 31.8862
10 310 266 44.4910 44.4910
11 240 274 -34.4072 34.4072
12 230 268 -37.5257 37.5257
13 260

𝑦 ̂_(𝑡+1)=𝛼𝑦_𝑡+(1−
𝛼)𝑦 ̂_𝑡
MFE MAE

8.34 39.80

MFE indicates whether the forecast is smaller

or larger than the actual data in general, but it
can mislead the manager into thinking that the
average error is low because positive and
negative forecast errors can offset each other.
So, the mean error is not a very useful
measure of forecast accuracy.

MAE can measure the deviation as it avoids

the problem of positive and negative forecast
errors offsetting each other.

MSE can measure the deviation, but it is more

sensitive to large errors as the squaring of
large errors can exaggerate their importance
for a time series.

MAPE provides relative or percentage error

measures, and it can also be used to make
comparisons for different time intervals (such
as comparing a method of forecasting monthly
gasoline sales to a method of forecasting
weekly sales) or to make comparisons across
different time series (such as monthly sales of
gasoline and monthly sales of oil filters).
Squared Forecast Error Absolute Percentage Error

177.78 5.13%
0.00 0.00%
4011.11 19.79%
4444.44 30.30%
1344.44 11.83%
1877.78 18.06%
2844.44 17.20%
2177.78 19.44%
1111.11 14.49%

MSE MAPE

1998.77 15.14%

Squared Forecast Error Absolute Percentage Error

64.00 3.08%
121.00 3.93%
3136.00 17.50%
5776.00 34.55%
2304.00 15.48%
2025.00 18.75%
2809.00 17.10%
2401.00 20.42%
961.00 13.48%

MSE MAPE

2177.44 16.03%

Squared Forecast Error Absolute Percentage Error

0.0000 0.00%
12100.0000 31.43%
1024.0000 13.91%
19.3600 1.69%
553.1904 8.40%
3459.3219 18.38%
2803.4060 24.07%
2269.7830 15.37%
1016.7303 13.29%
1979.4521 14.35%
1183.8536 14.34%
1408.1810 16.32%

MSE MAPE

2318.11 14.29%
 α 0.3
Week Actual Demand Exponential Smoothing (α=0.3)
1 31 31
2 34 31
3 33 32
4 35 32
5 37 33
6 36 34
7 38 35
8 40 36
9 40 37

𝑦 ̂_(𝑡+1)=𝛼𝑦_𝑡+(1−
10 41 38

𝛼)𝑦 ̂_𝑡

α 0.3
δ 0.2
Week Actual Demand (yt) ŷt Tt FITt (α=0.3, δ=0.2)
1 31 30 1 31.0000
2 34 31 1.00 32.0000
3 33 33 1.12 33.7200
4 35 34 1.08 34.5808
5 37 35 1.10 35.8085
6 36 36 1.17 37.3394
7 38 37 1.09 38.0307
8 40 38 1.09 39.1127
9 40 39 1.14 40.5234
10 41 40 1.11 41.4794

Time Series
45

43

41

39

37
Demand

35

33
 41

39

37

Demand
35

33

31

29

27

25
1 2 3 4 5 6 7 8

Week
 Forecast Error Absolute Forecast Error Squared Forecast Error
0.0000 0.0000 0.0000
3.0000 3.0000 9.0000
1.1000 1.1000 1.2100
2.7700 2.7700 7.6729
3.9390 3.9390 15.5157
1.7573 1.7573 3.0881
3.2301 3.2301 10.4336
4.2611 4.2611 18.1568
2.9828 2.9828 8.8968
3.0879 3.0879 9.5353

MFE MAE MSE

2.61 2.61 8.35

Forecast Error Absolute Forecast Error Squared Forecast Error

0.0000 0.0000 0.0000
2.0000 2.0000 4.0000
-0.7200 0.7200 0.5184
0.4192 0.4192 0.1757
1.1915 1.1915 1.4196
-1.3394 1.3394 1.7940
-0.0307 0.0307 0.0009
0.8873 0.8873 0.7873
-0.5234 0.5234 0.2739
-0.4794 0.4794 0.2299

MFE MAE MSE

0.14 0.76 0.92

Time Series Plots

Actual Demand
Exponential Smoothing (α=0.3)
FITt (α=0.3, δ=0.2)
Linear (FITt (α=0.3, δ=0.2))
 Actual Demand
Exponential Smoothing (α=0.3)
FITt (α=0.3, δ=0.2)
Linear (FITt (α=0.3, δ=0.2))

5 6 7 8 9 10

Week
Absolute Percentage Error
0.00%
8.82%
3.33%
7.91%
10.65%
4.88%
8.50%
10.65%
7.46%
7.53%

MAPE

6.97%

Absolute Percentage Error

0.00%
5.88%
2.18%
1.20%
3.22%
3.72%
0.08%
2.22%
1.31%
1.17%

MAPE

2.10%
 t/X Y x^2 XY Forecast Forecast Error Squared Forecast Error
1 120 1 120 114.79 5.21 27.19
2 110 4 220 109.86 0.14 0.02
3 100 9 300 104.93 -4.93 24.29
4 96 16 384 100.00 -4.00 16.00
5 94 25 470 95.07 -1.07 1.15
6 92 36 552 90.14 1.86 3.45
7 88 49 616 85.21 2.79 7.76
Total 28 700 140 2662
Mean 4 100 11.41

slope -4.928571 -4.928571

intercept 119.7143 119.7143

ŷ = 119.714 - 4.9286(t)
When t=8, ŷ = 119.714 - 4.9286(8) 150
ŷ = 80.29 100

Y
50
0
0
=MSE

Y vs t
150
100 f(x) = − 4.92857142857143 x + 119.714285714286
R² = 0.894924812030076
Y

50
0
0 1 2 3 4 5 6 7 8
t