Professional Documents
Culture Documents
Assistant Professor
Indian Institute of Information Technology
Introduction
Forecasting is essential for a number of planning decisions
Constraints
• Budget
• Space
• Resources
- Men , Equipment
A Forecast of Demand is an essential Input for Planning
METHOD OF FORECASTING
F P
Objective Subjective
Scientific Intuitive
Free from ‘BIAS’ Individual BIAS
Reproducible Non – Reproducible
Error Analysis Error Analysis
Possible Limited
OPINION POLLS
• Personal interviews
e.g. aggregation of opinion of sales
representatives to obtain sales forecast
of a region
• Knowledge base (experience)
• Subjective bias
• Questionnaire method
• questionnaire design
• choice of respondents
• obtaining respondents
• analysis and presentation of results (forecasting)
• Telephonic conversation
• Fast
• DELPHI
COMMONLY OBSERVED
“NORMAL” DEMAND PATTERNS
D Seasonal
D
Cyclic Pattern with
Growth
t t
ABNORMAL DEMAND
PATTERNS
DELPHI
Coordinator Expert 2
Expert n
A Statistical
• Mean summary
•Median can be given
1990 1995 2000 2005 2010 •Std. deviation at end of
year
each round
DELPHI (Contd.)
Round
1
Moving
Towards
Consensus Round
2
Round
3
DELPHI (Contd.)
Round
1
Moving
Towards
Divergent
View Points Round
2
Round
3
CHARACTERISTICS OF MOVING AVERAGES
Dt Dt
t t
Jan 199
Feb 202
Mar 199 200.00
Apr 208 203.00
May 212 206.33
Jun 194 204.66 202.33
Jul 214 205.66 207.83
Aug 220 208.33 210.83
Sep 219 216.66 213.13
Oct 234 223.33 217.46
Nov 219 223.00 218.63
Dec 233 227.66 225.13
K PERIOD MA = AVERAGE OF K MOST
RECENT OBSERVATIONS
220
6 Month
210 MA
200 6 Month
MA
190
J F M A M J J A S O N D J
Month
Weighted moving average
method:
In some conditions it may desirable to apply unequal
weights to the historical data
If more recent data are to be believed to be more
relevant to a forecast period, larger weights could be
applied
• Lets take an example
week Actual weight
data
1 85 .20
2 102 .30
3 110 .50
The forecast for the 4th week
F10 =85x.20 + 102x.30 + 110x.50= 102.6
EXPONENTIAL SMOOTHING
(Equivalence between & N :)
• = 2 / (N+1)
EXPONENTIAL SMOOTHING
F10 F9 ( A9 F9 )
0.1 F10 86.7 .1(110 86.7) 89.0
0.2 F10 88.4 .2(110 88.4) 92.7
0.3 F10 90.1 .3(110 90.1) 96.1
Exponential smoothing with
trend
While considering the short range forecast move
towards medium range forecast, seasonality and trend
becomes more important.
Incorporating a trend component into exponentially
smoothed forecast is called double exponential
smoothing because the estimate average and the
estimate for the trend are both smoothed.
α= the smoothing constant for the average
β= the smoothing constant for the trend
Model :
St = smoothed forecast
Tt = trend estimate period
At = actual demand in period t
t = the next time period
t- 1= the prior time period
FTt = forecast with trend in period t
α = smoothing constant for the average from 0 to 1
β= smoothing constant for the trend from 0 to 1.