Scribd
Upload
143 views4 pages

Forecasting Error Analysis

1. This document compares the forecast accuracy of 4 different forecasting methods: single exponential smoothing, weighted moving average, double exponential smoothing, and simple moving average. It provides the mean forecast error, mean absolute error, and mean absolute percentage error for each method based on historical sales data from 2003 to 2019. 2. Double exponential smoothing produced the lowest mean absolute percentage error of 9.25% while simple moving average produced the highest at 19.17%. Weighted moving average had the highest errors for individual periods, with a mean absolute percentage error of 64.02%. 3. In summary, double exponential smoothing provided the most accurate forecasts overall according to the error metrics, while simple moving average and weighted moving average tended to

Uploaded by

raymond balite
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
143 views4 pages

Forecasting Error Analysis

1. This document compares the forecast accuracy of 4 different forecasting methods: single exponential smoothing, weighted moving average, double exponential smoothing, and simple moving average. It provides the mean forecast error, mean absolute error, and mean absolute percentage error for each method based on historical sales data from 2003 to 2019. 2. Double exponential smoothing produced the lowest mean absolute percentage error of 9.25% while simple moving average produced the highest at 19.17%. Weighted moving average had the highest errors for individual periods, with a mean absolute percentage error of 64.02%. 3. In summary, double exponential smoothing provided the most accurate forecasts overall according to the error metrics, while simple moving average and weighted moving average tended to

Uploaded by

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

1.

Moving Average

Actual Absolute Percent


Periods Forecast Error
Value Error Error
2017 jan 1500 2018
feb 2200
mar 2700
apr 4200
may 7800
jun 5400 3680 1720 1720 31.85%
jul 4800 4460 340 340 7.08%
aug 5500 4980 520 520 9.45%
sep 6000 5540 460 460 7.67%
oct 7500 5900 1600 1600 21.33%
nov 4400 5840 -1440 1440 32.73%
dec 4545 5640 -1095 1095 24.09%
2018 jan 5589 2019
2105 7175 134.21%

Mean Forcast Error (MFE) 300.7143


Mean Absolute Error (MAE) 1025
Mean Absolute Percentage Error (MAPE) 19.17%

1. Single Exponential Smoothing


α = 0.3
Energy Absolute Percent
Year Forecast Error
Sales Error Error
2003 1192359
2004 1329893 1329893 0 0 0.00%
2005 1336189 1329893 6296 6296 0.47%
2006 1204297 1331782 -127485 127484.8 10.59%
2007 1353285 1293536 59748.64 59748.64 4.42%
2008 1466632 1311461 155171 155171 10.58%
2009 1728856 1358012 370843.7 370843.7 21.45%
2010 2027407 1469265 558141.6 558141.6 27.53%
2011 2200065 1636708 563357.1 563357.1 25.61%
2012 2400642 1805715 594927 594927 24.78%
2013 2258937 1984193 274743.9 274743.9 12.16%
2014 2403032 2066616 336415.7 336415.7 14.00%
2015 2548494 2167541 380953 380953 14.95%
2016 2818421 2281827 536594.1 536594.1 19.04%
2017 3292037 2442805 849231.9 849231.9 25.80%
2018 3518356 2697575 820781.3 820781.3 23.33%
2019 2943809
5379720 5634690 234.69%

Mean Forcast Error (MFE) 358648


Mean Absolute Error (MAE) 375646
Mean Absolute Percentage Error (MAPE) 15.65%
1. Weighted Moving Average
n=4
Actual Absolute Percent
Periods Forecast Error
Value Error Error
jan 5400
feb 4800
mar 5500
apr 6000
may 7500 5550 1950 1950 26.00%
jun 4400 6380 -1980 1980 45.00%
jul 4545 5760 -1215 1215 26.73%
aug 1500 5238 -3738 3738 249.20%
sep 2200 3593.5 -1393.5 1393.5 63.34%
oct 2700 2679 21 21 0.78%
nov 4200 2494.5 1705.5 1705.5 40.61%
dec 7800 3080 4720 4720 60.51%
jan 5140
70 16723 512.17%

Mean Forcast Error (MFE) 8.75


Mean Absolute Error (MAE) 2090.375
Mean Absolute Percentage Error (MAPE) 64.02%

1. Double Exponential Smoothing


α = 0.3 ß = 0.2
Energy Absolute Percent
Year Ln Tn Forecast Error
Sales Error Error
2003 1192359
2004 1329893 1329893 137534 0 0 0 0.00%
2005 1336189 1428056 129659.7 1467427 -131238 131238 9.82%
2006 1204297 1451690 108454.6 1557715 -353418 353418.3 29.35%
2007 1353285 1498087 96043.05 1560144 -206859 206859.4 15.29%
2008 1466632 1555880 88393.19 1594130 -127498 127497.7 8.69%
2009 1728856 1669648 93468.14 1644274 84582.44 84582.44 4.89%
2010 2027407 1842404 109325.6 1763116 264290.6 264290.6 13.04%
2011 2200065 2026230 124225.7 1951729 248335.8 248335.8 11.29%
2012 2400642 2225512 139236.9 2150456 250186.4 250186.4 10.42%
2013 2258937 2333005 132888.2 2364748 -105811 105811.5 4.68%
2014 2403032 2447035 129116.5 2465893 -62861.2 62861.24 2.62%
2015 2548494 2567854 127457.1 2576151 -27657.4 27657.41 1.09%
2016 2818421 2732244 134843.7 2695311 123109.7 123109.7 4.37%
2017 3292037 2994573 160340.6 2867088 424949.1 424949.1 12.91%
2018 3518356 3263946 182147.2 3154913 363442.8 363442.8 10.33%
2019 2412265 -24618.4 3446093
743553.2 2774240 138.78%

Mean Forcast Error (MFE) 49570.22


Mean Absolute Error (MAE) 184949.4
Mean Absolute Percentage Error (MAPE) 9.25%

You might also like