Embed
Email

Forecasting

Document Sample
Forecasting
Shared by: HC11111814942
Categories
Tags
Stats
views:
3
posted:
11/18/2011
language:
English
pages:
35
Forecasting

• Purpose is to forecast, not to explain the

historical pattern

• Models for forecasting may not make sense

as a description for ”physical” beaviour of

the time series

• Common sense and mathematics in a good

combination produces ”optimal” forecasts

• With time series regression models,

forecasting (prediction) is a natural step and

forecasting limits (intervals) can be

constructed

• With Classical decomposition, forecasting

may be done, but estimation of accuracy

lacks and no forecasting limits are produced

• Classical decomposition is usually

combined with Exponential smoothing

methods

Exponential smoothing

• Use the historical data to forecast the future

• Let different parts of the history have

different impact on the forecasts

• Forecast model is not developed from any

statistical theory

Single exponential smoothing

• Assume historical values y1,y2,…yT

• Assume data contains no trend, i.e.

yt   0   t

Forecasting scheme:





 T    yT  (1   )   T 1 ,

yT    T

ˆ



where  is a smoothing parameter

between 0 and 1

• The forecast procedure is a recursion

formula

• How shall we choose α?

• Where should we start, i.e. Which is the

initial value l0 ?

Use a part (usually half) of the historical data to

ˆ

estimate β0   0







Set l 0= ˆ

0





Update the estimates of β0 for the rest of the

historical data with the recursion formula





 l T which can be used to forecast yT+τ

Example: Sales of everyday commodities

Year Sales values

1985 151

1986 151

1987 147

150

1988 149

1989 146

1990 142









sales

145

1991 143

1992 145

1993 141 140



1994 143

1995 145 1985 1990 1995 2000

1996 138 year

1997 147

1998 151

1999 148

2000 148

Assume the model:





yt   0   t



Estimate β0 by calculating the mean value of the

first 8 observations of the series





ˆ

 0  (151 151 ...145)/8 146.75





ˆ

Set l8 =  0 =146.75

Assume first that the sales are very stable, i.e. during

the period the mean value β0 is assumed not to change





Set α to be relatively small. This means that the latest

observation plays a less role than the history in the

forecasts. Thumb rule: 0.05 Name c3 "FORE1" c4 "UPPE1" c5 "LOWE1"

MTB > SES 'Sales values';

SUBC> Weight 0.1;

SUBC> Initial 8;

SUBC> Forecasts 3;

SUBC> Fstore 'FORE1';

SUBC> Upper 'UPPE1';

SUBC> Lower 'LOWE1';

SUBC> Title "SES alpha=0.1".





Single Exponential Smoothing for Sales values





Data Sales values

Length 16









Smoothing Constant





Alpha 0.1

Accuracy Measures





MAPE 2.2378

MAD 3.2447

MSD 14.4781









Forecasts





Period Forecast Lower Upper

17 146.043 138.094 153.992

18 146.043 138.094 153.992

19 146.043 138.094 153.992

MINITAB uses smoothing

from 1st value!

Assume now that the sales are less stable, i.e. during the

period the mean value β0 is possibly changing





Set α to be relatively large. This means that the latest

observation becomes more important in the forecasts.





E.g. Set α=0.5 (A bit exaggerated)

Single Exponential Smoothing for Sales values





Data Sales values

Length 16





Smoothing Constant





Alpha 0.5





Accuracy Measures





MAPE 1.9924

MAD 2.8992

MSD 13.0928





Forecasts





Period Forecast Lower Upper

17 147.873 140.770 154.976

18 147.873 140.770 154.976

19 147.873 140.770 154.976

Slightly wider prediction intervals

We can also use some adaptive procedure to continuosly

evaluate the forecast ability and maybe change the

smoothing parameter over time

Alt. We can run the process with different alphas and

choose the one that performs best. This can be done with

the MINITAB procedure.

Single Exponential Smoothing for Sales values

---

Smoothing Constant SES optimal alpha

156 Variable

Actual

Alpha 0.567101 Fits

Forecasts

152 95.0% PI



Smoothing C onstant

A lpha 0.567101

Accuracy Measures



Sales values

148 Accuracy Measures

MAPE 1.7914

MAD 2.5940

MSD 12.1632

MAPE 1.7914 144

MAD 2.5940

MSD 12.1632 140







2 4 6 8 10 12 14 16 18

Forecasts Index







Period Forecast Lower Upper

17 148.013 141.658 154.369 Yet, wider prediction

18 148.013 141.658 154.369 intervals

19 148.013 141.658 154.369

Exponential smoothing for times series with trend

and/or seasonal variation





• Double exponential smoothing (one smoothing

parameter) for trend

• Holt’s method (two smoothing parameters) for

trend

• Multiplicative Winter’s method (three smoothing

parameters) for seasonal (and trend)

• Additive Winter’s method (three smoothing

parameters) for seasonal (and trend)

Example: Real Estate Price Index for Weekend

Cottages in Sweden

Year REPI_C

1993 226

Time Series Plot of REPI_C

1994 241

600

1995 239

1996 240 500



1997 268

REPI_C

1998 303 400





1999 336

300

2000 414

2001 472

200

2002 496 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

2003 505

2004 546

2005 591 Trend but no seasonal variation

Applying Holt’s method with MINITAB (denoted Double

exponential smoothing in Minitab)

2 smoothing

parameters, one for

level and one for trend.

Option to let Minitab

calculate optimal

parameters.









Smoothing parameters should

still be kept low (0.05,0.3)

Double Exponential Smoothing for REPI_C





Data REPI_C

Length 13

Double Exponential Smoothing Plot for REPI_C

Variable

700 Actual

Smoothing Constants Fits

Forecasts

600 95.0% PI



Alpha (level) 0.2 Smoothing Constants

Alpha (lev el) 0.2

500 Gamma (trend) 0.2

Gamma (trend) 0.2 REPI_C

Accuracy Measures

400 MAPE 9.78

MAD 30.15

Accuracy Measures MSD 1160.79

300





MAPE 9.78 200



MAD 30.15

100

MSD 1160.79 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Index



Forecasts





Period Forecast Lower Upper

14 611.411 537.537 685.286

15 646.167 570.753 721.581

Example: Quarterly sales data

year quarter sales

1991 1 124

1991 2 157

1991 3 163

1991 4 126 Time Series Plot of sales

200

1992 1 119

190

1992 2 163

180

1992 3 176

1992 4 127 170



1993 1 126 sales 160



1993 2 160 150



1993 3 181 140



1993 4 121 130



1994 1 131 120



1994 2 168 110

1994 3 189 Quarter Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3

Year 1991 1992 1993 1994 1995

1994 4 134

1995 1 133

1995 2 167

1995 3 195

1995 4 131

Applying Winter’s multiplicative method with MINITAB

3 smoothing parameters, one for level, one for trend an one for seasonal variation.

No option to calculate optimal parameters. Choices have do be based on visual

inspection of the times series

Winters' Method for sales

Multiplicative Method

Data sales Winters' Method Plot for sales

Multiplicative Method

Length 20

210 Variable

Actual

200 Fits

Smoothing Constants Forecasts

190 95.0% PI

Alpha (level) 0.2 180 Smoothing Constants

Alpha (lev el) 0.2

Gamma (trend) 0.2 170 Gamma (trend) 0.2



sales

Delta (seasonal) 0.2

Delta (seasonal) 0.2 160

Accuracy Measures

150 MAPE 2.6446

MAD 3.8808

Accuracy Measures 140 MSD 23.7076



MAPE 2.6446 130

MAD 3.8808 120

MSD 23.7076 Quarter Q3 Q3 Q3 Q3 Q3 Q3

Year 2008 2009 2010 2011 2012 2013



Forecasts

Period Forecast Lower Upper

Q3-2013 135.625 126.117 145.133

Q4-2013 174.430 164.773 184.087

Q1-2014 194.667 184.844 204.490

Q2-2014 136.933 126.928 146.939


Related docs
Other docs by HC11111814942
Ecological theology:
Views: 0  |  Downloads: 0
DECC Annual Report 2008-09: Appendices
Views: 0  |  Downloads: 0
Slide 1
Views: 0  |  Downloads: 0
LRPA-TV
Views: 0  |  Downloads: 0
PowerPoint Presentation
Views: 0  |  Downloads: 0
?????????? 10
Views: 2  |  Downloads: 0
Chart
Views: 0  |  Downloads: 0
PowerPoint-Pr�sentation
Views: 0  |  Downloads: 0
Slide 1
Views: 0  |  Downloads: 0
By registering with docstoc.com you agree to our
privacy policy

You are almost ready to download!

You are almost ready to download!