Upload
prudence-lambert
View
256
Download
2
Embed Size (px)
DESCRIPTION
Ch. 4: What is not covered? Trend Adjusted Exponential SmoothingTrend Adjusted Exponential Smoothing Tracking SignalTracking Signal By-Hand computation of anything done by ExcelBy-Hand computation of anything done by Excel –a and b using Least Squares –MAD, CE, Bias, MSE, MAPD, r and r 2 Multiple RegressionMultiple Regression
Citation preview
Chapter 4Chapter 4
ForecastingForecasting
Ch. 4: What is covered?Ch. 4: What is covered?• Moving AverageMoving Average• Weighted Moving AverageWeighted Moving Average• Exponential SmoothingExponential Smoothing• Trend ProjectionsTrend Projections• Seasonal Index/Adjusted-ForecastSeasonal Index/Adjusted-Forecast• MAD, CE, Bias, MSE, MAPDMAD, CE, Bias, MSE, MAPD• Linear Regression, r, rLinear Regression, r, r22
Ch. 4: What is Ch. 4: What is notnot covered? covered?• Trend Adjusted Exponential SmoothingTrend Adjusted Exponential Smoothing• Tracking SignalTracking Signal• By-Hand computation of anything done by By-Hand computation of anything done by
ExcelExcel– a and b using Least Squaresa and b using Least Squares– MAD, CE, Bias, MSE, MAPD, r and rMAD, CE, Bias, MSE, MAPD, r and r22
• Multiple RegressionMultiple Regression
ForecastingForecastingPredicting future eventsPredicting future eventsUsually demand behavior Usually demand behavior
over a time frameover a time frameQualitative methodsQualitative methods
Based on subjective methodsBased on subjective methodsQuantitative methodsQuantitative methods
Based on mathematical formulasBased on mathematical formulas
Demand BehaviorDemand BehaviorTrendTrend
gradual, long-term up or down gradual, long-term up or down movementmovement
CycleCycle up & down movement repeating over up & down movement repeating over
long time framelong time frameSeasonal patternSeasonal pattern
periodic oscillation in demand which periodic oscillation in demand which repeatsrepeats
Random movements follow no patternRandom movements follow no pattern
Forms of Forecast MovementForms of Forecast Movement
TimeTime(a) Trend(a) Trend
TimeTime(d) Trend with seasonal pattern(d) Trend with seasonal pattern
TimeTime(c) Seasonal pattern(c) Seasonal pattern
TimeTime(b) Cycle(b) Cycle
Dem
and
Dem
and
Dem
and
Dem
and
Dem
and
Dem
and
Dem
and
Dem
and
Random Random movementmovement
Moving AverageMoving Average
MAMAnn = =
nn
ii = 1= 1 DDii
nnwherewhere
nn ==number of periods number of periods in the moving in the moving averageaverage
DDii ==demand in period demand in period ii
Average several Average several periods of dataperiods of data
Dampen, smooth out Dampen, smooth out changeschanges
Use when demand is Use when demand is stable with no trend stable with no trend or seasonal patternor seasonal pattern
JanJan 120120FebFeb 9090MarMar 100100AprApr 7575MayMay 110110JuneJune 5050JulyJuly 7575AugAug 130130SeptSept 110110OctOct 9090
ORDERSORDERSMONTHMONTH PER MONTHPER MONTH
MAMA33 = =
33
ii = 1= 1 DDii
33
== 90 + 110 + 13090 + 110 + 13033
= 110 orders for Nov= 110 orders for Nov
Simple Moving AverageSimple Moving Average
JanJan 120120 ––FebFeb 9090 – –MarMar 100100 – –AprApr 7575 103.3103.3MayMay 110110 88.388.3JuneJune 5050 95.095.0JulyJuly 7575 78.378.3AugAug 130130 78.378.3SeptSept 110110 85.085.0OctOct 9090 105.0105.0NovNov – – 110.0110.0
ORDERSORDERS THREE-MONTHTHREE-MONTHMONTHMONTH PER MONTHPER MONTH MOVING AVERAGEMOVING AVERAGE
Simple Moving AverageSimple Moving Average
JanJan 120120 ––FebFeb 9090 – –MarMar 100100 – –AprApr 7575 103.3103.3MayMay 110110 88.388.3JuneJune 5050 95.095.0JulyJuly 7575 78.378.3AugAug 130130 78.378.3SeptSept 110110 85.085.0OctOct 9090 105.0105.0NovNov – – 110.0110.0
ORDERSORDERS THREE-MONTHTHREE-MONTHMONTHMONTH PER MONTHPER MONTH MOVING AVERAGEMOVING AVERAGE
MAMA55 = =
55
ii = 1= 1 DDii
55
== 90 + 110 + 130 + 75 + 5090 + 110 + 130 + 75 + 5055
= 91 orders for Nov= 91 orders for Nov
Simple Moving AverageSimple Moving Average
Simple Moving AverageSimple Moving Average
JanJan 120120 –– – –FebFeb 9090 – – – –MarMar 100100 – – – –AprApr 7575 103.3103.3 – –MayMay 110110 88.388.3 – –JuneJune 5050 95.095.0 99.099.0JulyJuly 7575 78.378.3 85.085.0AugAug 130130 78.378.3 82.082.0SeptSept 110110 85.085.0 88.088.0OctOct 9090 105.0105.0 95.095.0NovNov – – 110.0110.0 91.091.0
ORDERSORDERS THREE-MONTHTHREE-MONTH FIVE-MONTHFIVE-MONTHMONTHMONTH PER MONTHPER MONTH MOVING AVERAGEMOVING AVERAGE MOVING AVERAGEMOVING AVERAGE
Smoothing EffectsSmoothing Effects150 150 –
125 125 –
100 100 –
75 75 –
50 50 –
25 25 –
0 0 –| | | | | | | | | | |
JanJan FebFeb MarMar AprApr MayMay JuneJune JulyJuly AugAug SeptSept OctOct NovNov
5-month5-month
3-month3-month
ActualActual
Ord
ers
Ord
ers
MonthMonth
Weighted Moving AverageWeighted Moving Average
WMAWMAnn = = ii = 1 = 1 WWii D Dii
wherewhere
WWii = the weight for period = the weight for period ii, , between 0 and 100 between 0 and 100 percentpercent
WWii = 1.00= 1.00
Adjusts Adjusts moving moving average average method to method to more closely more closely reflect data reflect data fluctuationsfluctuations
Weighted Moving Weighted Moving Average ExampleAverage Example
MONTH MONTH WEIGHT WEIGHT DATADATAAugustAugust 17%17% 130130SeptemberSeptember 33%33% 110110OctoberOctober 50%50% 9090
Weighted Moving Weighted Moving Average ExampleAverage Example
MONTH MONTH WEIGHT WEIGHT DATADATAAugustAugust 17%17% 130130SeptemberSeptember 33%33% 110110OctoberOctober 50%50% 9090
November forecastNovember forecast WMAWMA33 = = 33
ii = 1 = 1 WWii D Dii
= (0.50)(90) + (0.33)(110) + (0.17)(130)= (0.50)(90) + (0.33)(110) + (0.17)(130)
= 103.4 orders= 103.4 orders
FFt t +1 +1 = = DDtt + (1 - + (1 - ))FFtt
wherewhereFFt t +1+1 = =forecast for next forecast for next periodperiodDDtt ==actual demand actual demand for present periodfor present periodFFtt ==previously previously determined forecast determined forecast for present periodfor present period
== weighting factor, weighting factor, smoothing constantsmoothing constant
Averaging method Averaging method Weights most Weights most
recent data more recent data more stronglystrongly
Reacts more to Reacts more to recent changesrecent changes
Widely used, Widely used, accurate methodaccurate method
Exponential SmoothingExponential Smoothing
Effect of Smoothing Effect of Smoothing ConstantConstant0.0 0.0 1.0 1.0
If If = 0.20, then = 0.20, then FFt t +1 +1 = 0.20= 0.20DDtt + 0.80 + 0.80 FFtt
If If = 0, then = 0, then FFtt +1 +1 = 0= 0DDtt + 1 + 1 FFtt 0 = 0 = FFtt
Forecast does not reflect recent dataForecast does not reflect recent data
If If = 1, then = 1, then FFt t +1 +1 = 1= 1DDtt + 0 + 0 FFtt ==DDtt Forecast based only on most recent dataForecast based only on most recent data
PERIODPERIOD MONTHMONTH DEMANDDEMAND
11 JanJan 373722 FebFeb 404033 MarMar 414144 AprApr 373755 May May 454566 JunJun 505077 Jul Jul 434388 Aug Aug 474799 Sep Sep 5656
1010 OctOct 52521111 NovNov 55551212 Dec Dec 5454
FF22 = = DD11 + (1 - + (1 - ))FF11
= (0.30)(37) + (0.70)(37)= (0.30)(37) + (0.70)(37)= 37= 37
FF33 = = DD22 + (1 - + (1 - ))FF22
= (0.30)(40) + (0.70)(37)= (0.30)(40) + (0.70)(37)= 37.9= 37.9
FF1313 = = DD1212 + (1 - + (1 - ))FF1212
= (0.30)(54) + (0.70)(50.84)= (0.30)(54) + (0.70)(50.84)= 51.79= 51.79
Exponential SmoothingExponential Smoothing
FORECAST, FORECAST, FFtt + 1 + 1
PERIODPERIOD MONTHMONTH DEMANDDEMAND (( = 0.3) = 0.3)
11 JanJan 3737 ––22 FebFeb 4040 37.0037.0033 MarMar 4141 37.9037.9044 AprApr 3737 38.8338.8355 May May 4545 38.2838.2866 JunJun 5050 40.2940.2977 Jul Jul 4343 43.2043.2088 Aug Aug 4747 43.1443.1499 Sep Sep 5656 44.3044.30
1010 OctOct 5252 47.8147.811111 NovNov 5555 49.0649.061212 Dec Dec 5454 50.8450.841313 JanJan –– 51.7951.79
Exponential SmoothingExponential Smoothing
FORECAST, FORECAST, FFtt + 1 + 1
PERIODPERIOD MONTHMONTH DEMANDDEMAND (( = 0.3) = 0.3) (( = 0.5) = 0.5)
11 JanJan 3737 –– ––22 FebFeb 4040 37.0037.00 37.0037.0033 MarMar 4141 37.9037.90 38.5038.5044 AprApr 3737 38.8338.83 39.7539.7555 May May 4545 38.2838.28 38.3738.3766 JunJun 5050 40.2940.29 41.6841.6877 Jul Jul 4343 43.2043.20 45.8445.8488 Aug Aug 4747 43.1443.14 44.4244.4299 Sep Sep 5656 44.3044.30 45.7145.71
1010 OctOct 5252 47.8147.81 50.8550.851111 NovNov 5555 49.0649.06 51.4251.421212 Dec Dec 5454 50.8450.84 53.2153.211313 JanJan –– 51.7951.79 53.6153.61
Exponential SmoothingExponential Smoothing
70 70 –
60 60 –
50 50 –
40 40 –
30 30 –
20 20 –
1010 –
0 0 –| | | | | | | | | | | | |11 22 33 44 55 66 77 88 99 1010 1111 1212 1313
ActualActual
Ord
ers
Ord
ers
MonthMonth
Exponential Smoothing Exponential Smoothing ForecastsForecasts
70 70 –
60 60 –
50 50 –
40 40 –
30 30 –
20 20 –
1010 –
0 0 –| | | | | | | | | | | | |11 22 33 44 55 66 77 88 99 1010 1111 1212 1313
ActualActual
Ord
ers
Ord
ers
MonthMonth
= 0.30= 0.30
Exponential Smoothing Exponential Smoothing ForecastsForecasts
70 70 –
60 60 –
50 50 –
40 40 –
30 30 –
20 20 –
1010 –
0 0 –| | | | | | | | | | | | |11 22 33 44 55 66 77 88 99 1010 1111 1212 1313
= 0.50= 0.50ActualActual
Ord
ers
Ord
ers
MonthMonth
= 0.30= 0.30
Exponential Smoothing Exponential Smoothing ForecastsForecasts
Mean Absolute Mean Absolute Deviation (MAD)Deviation (MAD)
wherewhere tt = the period number= the period number
DDtt = demand in period = demand in period tt FFtt = the forecast for period = the forecast for period tt
nn = the total number of periods= the total number of periods= the absolute value= the absolute value
DDtt - - FFtt nnMAD =MAD =
MAD ExampleMAD Example11 3737 37.0037.0022 4040 37.0037.0033 4141 37.9037.9044 3737 38.8338.8355 4545 38.2838.2866 5050 40.2940.2977 4343 43.2043.2088 4747 43.1443.1499 5656 44.3044.30
1010 5252 47.8147.811111 5555 49.0649.061212 5454 50.8450.84
557557
PERIODPERIOD DEMAND, DEMAND, DDtt FFtt ( ( =0.3) =0.3)
MAD ExampleMAD Example11 3737 37.0037.00 –– ––22 4040 37.0037.00 3.003.00 3.003.0033 4141 37.9037.90 3.103.10 3.103.1044 3737 38.8338.83 -1.83-1.83 1.831.8355 4545 38.2838.28 6.726.72 6.726.7266 5050 40.2940.29 9.699.69 9.699.6977 4343 43.2043.20 -0.20-0.20 0.200.2088 4747 43.1443.14 3.863.86 3.863.8699 5656 44.3044.30 11.7011.70 11.7011.70
1010 5252 47.8147.81 4.194.19 4.194.191111 5555 49.0649.06 5.945.94 5.945.941212 5454 50.8450.84 3.153.15 3.153.15
557557 49.3149.31 53.3953.39
PERIODPERIOD DEMAND, DEMAND, DDtt FFtt ( ( =0.3) =0.3) ((DDtt - - FFtt)) | |DDtt - - FFtt||
MAD ExampleMAD Example11 3737 37.0037.00 –– ––22 4040 37.0037.00 3.003.00 3.003.0033 4141 37.9037.90 3.103.10 3.103.1044 3737 38.8338.83 -1.83-1.83 1.831.8355 4545 38.2838.28 6.726.72 6.726.7266 5050 40.2940.29 9.699.69 9.699.6977 4343 43.2043.20 -0.20-0.20 0.200.2088 4747 43.1443.14 3.863.86 3.863.8699 5656 44.3044.30 11.7011.70 11.7011.70
1010 5252 47.8147.81 4.194.19 4.194.191111 5555 49.0649.06 5.945.94 5.945.941212 5454 50.8450.84 3.153.15 3.153.15
557557 49.3149.31 53.3953.39
PERIODPERIOD DEMAND, DEMAND, DDtt FFtt ( ( =0.3) =0.3) ((DDtt - - FFtt)) | |DDtt - - FFtt||
Dt - Ft nMAD =
=
= 4.45
53.3912
Other Accuracy MeasuresOther Accuracy Measures
Average error = biasAverage error = bias
E =E =eett
nnMean squared error =Mean squared error =
E =E =ee22
tt
nn
yy = = aa + + bxbx
wherewhereaa == intercept (at period 0)intercept (at period 0)bb == slope of the lineslope of the linexx == the time periodthe time periodyy == forecast for demand for period forecast for demand for period xx
Linear Trend LineLinear Trend Line
yy = = aa + + bxbx
wherewhereaa == intercept (at period 0)intercept (at period 0)bb == slope of the lineslope of the linexx == the time periodthe time periodyy == forecast for demand for period forecast for demand for period xx
b =
a = y - b x
wheren = number of periods
x = = mean of the x values
y = = mean of the y values
xy - nxy
x2 - nx2
xn
yn
Linear Trend LineLinear Trend Line
xx(PERIOD)(PERIOD) yy(DEMAND)(DEMAND)
11 373722 404033 414144 373755 454566 505077 434388 474799 5656
1010 52521111 55551212 5454
7878 557557
Least Squares ExampleLeast Squares Example
xx(PERIOD)(PERIOD) yy(DEMAND)(DEMAND) xyxy xx22
11 3737 3737 1122 4040 8080 4433 4141 123123 9944 3737 148148 161655 4545 225225 252566 5050 300300 363677 4343 301301 494988 4747 376376 646499 5656 504504 8181
1010 5252 520520 1001001111 5555 605605 1211211212 5454 648648 144144
7878 557557 38673867 650650
Least Squares ExampleLeast Squares Example
Least Squares ExampleLeast Squares Examplexx(PERIOD)(PERIOD) yy(DEMAND)(DEMAND) xyxy xx22
11 3737 3737 1122 4040 8080 4433 4141 123123 9944 3737 148148 161655 4545 225225 252566 5050 300300 363677 4343 301301 494988 4747 376376 646499 5656 504504 8181
1010 5252 520520 1001001111 5555 605605 1211211212 5454 648648 144144
7878 557557 38673867 650650
x = = 6.5
y = = 46.42
b =
=
= 1.72a = y - bx
= 46.42 - (1.72)(6.5)= 35.2
3867 - (12)(6.5)(46.42)650 - 12(6.5)2
xy - nxyx2 - nx2
781255712
Least Squares ExampleLeast Squares Examplexx(PERIOD)(PERIOD) yy(DEMAND)(DEMAND) xyxy xx22
11 7373 3737 1122 4040 8080 4433 4141 123123 9944 3737 148148 161655 4545 225225 252566 5050 300300 363677 4343 301301 494988 4747 376376 646499 5656 504504 8181
1010 5252 520520 1001001111 5555 605605 1211211212 5454 648648 144144
7878 557557 38673867 650650
x = = 6.5
y = = 46.42
b =
=
= 1.72a = y - bx
= 46.42 - (1.72)(6.5)= 35.2
3867 - (12)(6.5)(46.42)650 - 12(6.5)2
xy - nxyx2 - nx2
781255712
Linear trend liney = 35.2 + 1.72x
Least Squares ExampleLeast Squares Examplexx(PERIOD)(PERIOD) yy(DEMAND)(DEMAND) xyxy xx22
11 7373 3737 1122 4040 8080 4433 4141 123123 9944 3737 148148 161655 4545 225225 252566 5050 300300 363677 4343 301301 494988 4747 376376 646499 5656 504504 8181
1010 5252 520520 1001001111 5555 605605 1211211212 5454 648648 144144
7878 557557 38673867 650650
Linear trend liney = 35.2 + 1.72x
Forecast for period 13y = 35.2 + 1.72(13)
y = 57.56 units
Linear Trend LineLinear Trend Line70 70 –
60 60 –
50 50 –
40 40 –
30 30 –
20 20 –
1010 –
0 0 –| | | | | | | | | | | | |11 22 33 44 55 66 77 88 99 1010 1111 1212 1313
ActualActual
Dem
and
Dem
and
PeriodPeriod
Linear Trend LineLinear Trend Line70 70 –
60 60 –
50 50 –
40 40 –
30 30 –
20 20 –
1010 –
0 0 –| | | | | | | | | | | | |11 22 33 44 55 66 77 88 99 1010 1111 1212 1313
ActualActual
Dem
and
Dem
and
PeriodPeriod
Linear trend lineLinear trend line
Seasonal AdjustmentsSeasonal Adjustments
Repetitive increase/ Repetitive increase/ decrease in demanddecrease in demand
Use seasonal factor Use seasonal factor to adjust forecastto adjust forecast
Seasonal factor = Seasonal factor = SSii = =DDii
DD
Seasonal AdjustmentSeasonal Adjustment
1999 12.61999 12.6 8.68.6 6.36.3 17.517.5 45.045.02000 14.12000 14.1 10.310.3 7.57.5 18.218.2 50.150.12001 15.32001 15.3 10.610.6 8.18.1 19.619.6 53.653.6Total 42.0Total 42.0 29.529.5 21.921.9 55.355.3 148.7148.7
DEMAND (1000’S PER QUARTER)DEMAND (1000’S PER QUARTER)YEARYEAR 11 22 33 44 TotalTotal
Seasonal AdjustmentSeasonal Adjustment
1999 12.61999 12.6 8.68.6 6.36.3 17.517.5 45.045.02000 14.12000 14.1 10.310.3 7.57.5 18.218.2 50.150.12001 15.32001 15.3 10.610.6 8.18.1 19.619.6 53.653.6Total 42.0Total 42.0 29.529.5 21.921.9 55.355.3 148.7148.7
DEMAND (1000’S PER QUARTER)DEMAND (1000’S PER QUARTER)YEARYEAR 11 22 33 44 TotalTotal
SS11 = = = 0.28 = = = 0.28 DD11
DD42.042.0
148.7148.7
SS22 = = = 0.20 = = = 0.20 DD22
DD29.529.5
148.7148.7 SS44 = = = 0.37 = = = 0.37 DD44
DD55.355.3
148.7148.7
SS33 = = = 0.15 = = = 0.15 DD33
DD21.921.9
148.7148.7
Seasonal AdjustmentSeasonal Adjustment
1999 12.61999 12.6 8.68.6 6.36.3 17.517.5 45.045.02000 14.12000 14.1 10.310.3 7.57.5 18.218.2 50.150.12001 15.32001 15.3 10.610.6 8.18.1 19.619.6 53.653.6Total 42.0Total 42.0 29.529.5 21.921.9 55.355.3 148.7148.7
DEMAND (1000’S PER QUARTER)DEMAND (1000’S PER QUARTER)YEARYEAR 11 22 33 44 TotalTotal
SSii 0.280.28 0.200.20 0.150.15 0.370.37
Seasonal AdjustmentSeasonal Adjustment
1999 12.61999 12.6 8.68.6 6.36.3 17.517.5 45.045.02000 14.12000 14.1 10.310.3 7.57.5 18.218.2 50.150.12001 15.32001 15.3 10.610.6 8.18.1 19.619.6 53.653.6Total 42.0Total 42.0 29.529.5 21.921.9 55.355.3 148.7148.7
DEMAND (1000’S PER QUARTER)DEMAND (1000’S PER QUARTER)YEARYEAR 11 22 33 44 TotalTotal
SSii 0.280.28 0.200.20 0.150.15 0.370.37
yy = 40.97 + 4.30= 40.97 + 4.30xx= 40.97 + 4.30(4)= 40.97 + 4.30(4)= 58.17= 58.17
For 2002For 2002
Seasonal AdjustmentSeasonal Adjustment
SFSF11 = (= (SS11) () (FF44)) SFSF33 = (= (SS33) () (FF44) ) = (0.28)(58.17) = 16.28= (0.28)(58.17) = 16.28 = (0.15)(58.17) = 8.73= (0.15)(58.17) = 8.73
SFSF22 = (= (SS22) () (FF44)) SFSF44 = (= (SS44) () (FF44) ) = (0.20)(58.17) = 11.63= (0.20)(58.17) = 11.63 = (0.37)(58.17) = 21.53= (0.37)(58.17) = 21.53
1999 12.61999 12.6 8.68.6 6.36.3 17.517.5 45.045.02000 14.12000 14.1 10.310.3 7.57.5 18.218.2 50.150.12001 15.32001 15.3 10.610.6 8.18.1 19.619.6 53.653.6Total 42.0Total 42.0 29.529.5 21.921.9 55.355.3 148.7148.7
DEMAND (1000’S PER QUARTER)DEMAND (1000’S PER QUARTER)YEARYEAR 11 22 33 44 TotalTotal
SSii 0.280.28 0.200.20 0.150.15 0.370.37
yy = 40.97 + 4.30= 40.97 + 4.30xx= 40.97 + 4.30(4)= 40.97 + 4.30(4)= 58.17= 58.17
For 2002For 2002
Causal Modeling with Causal Modeling with Linear RegressionLinear Regression
Study relationship between two Study relationship between two or more variablesor more variables
Dependent variable Dependent variable yy depends depends on independent variable on independent variable xx
yy = = aa + + bxbx
CorrelationCorrelationCorrelation, Correlation, rr
Measure of strength of relationshipMeasure of strength of relationship Varies between -1.00 and +1.00Varies between -1.00 and +1.00 1.00 => an increase in the 1.00 => an increase in the
independent variable results in a independent variable results in a linear increase in the dependentlinear increase in the dependent
-1.00 -1.00 => an increase in the independent => an increase in the independent variable results in a linear decrease in the variable results in a linear decrease in the dependentdependent
0.0 => there does not seem to be a linear 0.0 => there does not seem to be a linear relationship between therelationship between the
Coefficient of Coefficient of DeterminationDetermination
• Coefficient of determination, rCoefficient of determination, r22
– Percentage of variation in dependent Percentage of variation in dependent variable resulting from changes in the variable resulting from changes in the independent variableindependent variable