Smoothing Methods
Smoothing Methods
Smoothing Methods
Marzena Narodzonek-Karpowska Prof. Dr. W. Toporowski Institut fr Marketing & Handel Abteilung Handel
What Is Forecasting?
Process of predicting a future event Underlying basis of all business decisions Youll never get it right. But, you can always get it less wrong.
Realities of Forecasting
Forecasts are seldom perfect Most forecasting methods assume that there is some underlying stability in the system and future will be like the past (causal factors will be the same). Accuracy decreases with length of forecast
Forecasting Methods
Qualitative Methods are subjective in nature since they rely on human judgment and opinion. Used when situation is vague & little data exist New products New technology Involve intuition, experience Quantitative Methods use mathematical or simulation models based on historical demand or relationships between variables. Used when situation is stable & historical data exist Existing products Current technology Involve mathematical techniques
Forecasting Methods
Forecasting Methods Quantitative Qualitative
Causal
Time Series Trend Projection Trend Projection Adjusted for Seasonal Influence
Smoothing
Linear
Quadratic
Exponential
Trend Component
The trend component accounts for the gradual shifting of the time series to relatively higher or lower values over a long period of time. Trend is usually the result of long-term factors such as changes in the population, demographics, technology, or consumer preferences
Sales
d rd tren pwa
Time
Cyclical Component
Any regular pattern of sequences of values above and below the trend line lasting more than one year can be attributed to the cyclical component. Usually, this component is due to multiyear cyclical movements in the economy. 1 Cycle Sales
Year
Seasonal Component
The seasonal component accounts for regular patterns of variability within certain time periods, such as a year. The variability does not always correspond with the seasons of the year (i.e. winter, spring, summer, fall). There can be, for example, within-week or within-day seasonal behavior. Sales
Summer Winter Spring Fall
Time (Quarterly)
Irregular Component
The irregular component is caused by short-term, unanticipated and non-recurring factors that affect the values of the time series. This component is the residual, or catch-all, factor that accounts for unexpected data values. It is unpredictable.
Trend
Seasonal
Cyclical
Irregular
Trend
1 2 3 4 5 6 7 8 9
Seasonal
10 11 12 13
Year
Smoothing Methods
In cases in which the time series is fairly stable and has no significant trend, seasonal, or cyclical effects, one can use smoothing methods to average out the irregular component of the time series. Common smoothing methods are: Moving Averages Weighted Moving Averages Centered Moving Average Exponential Smoothing
Moving Averages
Let us forecast sales for 2007 using a 3-period moving average. 2002 2003 2004 2005 2006 2007 4 6 5 3 7 ?
Moving Averages
Time
2002 1995 0032003 1996 2004 1997 2005 1998 2006 1999 2007 2000
Response Yi 4 6 5 3 7 NA
Sales 8 6
Forecast
4 2 95 96 97 Year 98 99 00
Actual
Demand
Time
WMA =
2x 12
Demand
Time
Actual sales
25 20 15 10 5 0
Ju l Au g Se p O ct No v Fe b M ar A pr M ay Ju n D ec Ja n
Moving average
Month
Ft +1 = Ft + ( y t Ft )
or: or
Ft +1 = y t + (1 )Ft
where: Ft+1= forecast value for period t + 1 yt = actual value for period t Ft = forecast value for period t = alpha (smoothing constant)
10
11
12
Ft +1 = y t + (1 )Ft
The smoothed value in this case is generally a little low, since the trend is upward sloping and the weighting factor is only 0.2
50 40
Sales
30 20 10 0 1 2 3 4 5 6 7 8 9 10
Quarter
Sales
Smoothed
Tt = (C t C t 1 ) + (1 )Tt 1
Ft +1 = C t + Tt
yt = actual value in time t = constant-process smoothing constant = trend-smoothing constant Ct = smoothed constant-process value for period t Tt = smoothed trend value for period t Ft+1= forecast value for period t + 1 t = current time period
Small Large
Demand
Time
Week 6 7 8 9 10
n=2 Yt Ft
#NV 112,5 120 122,5 122,5 122,5 125 122,5 112,5 120
110 115 125 120 125 120 130 115 110 130
n=3 Yt Ft
110 115 125 120 125 120 130 115 110 130
#NV #NV 116,6667 3,333333 11,11111 120 5 25 123,3333 -3,33333 11,11111 121,6667 8,333333 69,44444 125 -10 100 121,6667 -11,6667 136,1111 118,3333 11,66667 136,1111 118,3333 MSE 69,84127
Yt
110 115 125 120 125 120 130 115 110 130
=0.1 Ft
#NV 110 5 25 110,5 14,5 210,25 111,95 8,05 64,8025 112,755 12,245 149,94 113,9795 6,0205 36,24642 114,5816 15,41845 237,7286 116,1234 -1,1234 1,262016 116,0111 -6,01106 36,13279 115,4099 14,59005 212,8696 MSE 108,248
=0.2 Yt
110 115 125 120 125 120 130 115 110 130
Ft
#NV 110 5 111 14 113,8 6,2 115,04 9,96 117,032 2,968 117,6256 12,3744 120,1005 -5,10048 119,0804 -9,08038 117,2643 12,73569 MSE
Since the three period moving average technique (MA3) provides to lowest MSE value, this is the best smoothing technique to use for forecasting these Sales data in our example.
References
www.lsv.uni-saarland.de/Seminar/LMIR_WS0506/LM4IR_slides/ Alejandro_Figuera_Smoothing_Methods_for_LM_in_IR.ppt www.forecast.umkc.edu/ftppub/formula_pages/ch04lec.ppt www.angelfire.com/empire2/qnt531/workshop4.ppt www.swlearning.com/quant/asw/sbe_8e/powerpoint/ch18.ppt www.coursesite.cl.uh.edu/BPA/revere/DSCI5431/gradqntch_16&14.ppt www.cs.wright.edu/~rhill/EGR_702/Lecture%203.ppt www.lehigh.edu/~rhs2/ie409/BasicForecastingMethods.ppt www.cob.sjsu.edu/hibsho_a/ballou08aconcforecast.PPT www.mgtclass.mgt.unm.edu/MIDS/Kraye/Mgt%20300/Slide%20Show%2 0Week%20%232%20MGT%20300%20Forecasting.ppt www.infohost.nmt.edu/~toshi/EMGT%20501_files/Lectures/EMGT50113t hweek.ppt www.docp.wright.edu/bvi/classes/MBA780/MBA%20Forecasting-1.ppt www.cbtpdc.tamu-commerce.edu/gbus302/gbus302/ba578/Chapter14.ppt www.mgtclass.mgt.unm.edu/Weber/Mgt%20720/Unit%201/4.0_%20intro %20to%20forecasting.ppt