PRESENTATION ON FORECASTING

MODELS

12/8/21 NDIM 1
 What is forecasting?
• Forecasting is a basic tool which helps in managerial
decision making.
• For example in inventory control,marketing strategy
formulation,financial planning and production
planning
• It is of two types:
• Qualitative models
• Quantitative models

12/8/21 NDIM 2
 Forecasting
techniques

Qualitative Quantitative
models models

Delphi
method Time series Casual
models models
Sales force Moving Single
composite averages regression
Exponential
Consumer Multiple
smoothing
panel survey regression
Trend
projection

12/8/21 NDIM 3
 Qualitative models of forecasting
1.Delphi method
• This method is interactive group process and it
employs a group of experts to obtain the forecasts.
• The experts are not known to each other and their
interaction takes place through a coordinator
• Each individual forecast of one expert is given to thr
other expert and finally by reviewing they come to a
consensus

12/8/21 NDIM 4
2.Sales force composite
• All the members comprising the sales force of a
company are asked to estimate the likely sales in
their respective areas
• Estimates are reviewed to ensure that they are
realistic.
• Finally the estimates are combined at
district,regional and national level to obtain the
overall forecast

12/8/21 NDIM 5
3. Consumer panel survey
• In this method a consumer panel is made and are
questioned about their purchase plans.
• This method is based on strong assumption that the
consumers on the panel are truly representative of
the ultimate purchasers.

12/8/21 NDIM 6
Quantitative models of forecasting
1. Time series models-
These models use time-based data.
• A time series is a collection of readings belonging to
different time periods,equally spaced(may be
months,years or weeks)
• Prediction is done only on the basis on the past
data.

12/8/21 NDIM 7
1.1 Moving averages
• Ex-The demand for an item is observed for 15
months and recorded below:

Month Demand Month Demand

12/8/21 NDIM 8
Calculate
(i) 3-monthly
(ii) 4-monthly moving averages
What is the forecast for month 16 for each one?

12/8/21 NDIM 9
 Solution..
• 3 monthly moving average-immediately last
three months are considered and given
weights of 1/3 each and remaining value
weights zero.
• 4 monthly moving average- by taking
preceding 4 months average.

12/8/21 NDIM 10
Month Demand(y) 3-monthly 4-monthly 3-monthly
moving avg moving avg weighted
moving avg

1 280
2 288
3 266
4 295 278 275.7
5 302 283 282.3 284.2
6 310 287.7 287.8 293.7
7 303 302.3 293.3 304.8
8 328 305 302.5 305.2
9 309 313.7 310.8 316.7
10 315 313.3 312.5 314.3
11 320 317.3 313.8 315.2
12 332 314.7 318 316.5
13 310 322.3 319 325.2
14 308 320.7 319.3 319
15 320 316.7 317.5 312.7
16 12/8/21
312.7 NDIM
317.5 314.3 11
1.2 Exponential Smoothing
• Ex-The demand for a particular item during the 10 months
of a year is as given below.The manager is considering how
well the exponential smoothing serves as an appropriate
technique in forecasting the demand of this item.She is
testing three values of the smoothing constant α=0.2, α=0.5
and α=0.8
• You are required to
(a)Calculate forecasted values using each of the given α
values,assuming the initial forecast as 208,and
(b)Calculate Mean Absolute Deviation for each of these series
of estimate and suggest which of them is most appropriate.

12/8/21 NDIM 12
 Mont 1 2 3 4 5 6 7 8 9 10
h
Dem 213 201 198 207 220 232 210 217 212 225
and

Solution:-
F t+1=Ft+α (Yt-Ft)
so we will calculate these values for each value of
alpha
MAD=Sum of all (Yt-Ft) /n
We will find that of all the cases MAD is minimum
for α=0.2,therefore this smoothing constant is most
preferable of all the three cases.

12/8/21 NDIM 13
Month Demand α=0.2 α=0.5 α=0.8
(t) Yt Ft Yt –Ft Ft Yt –Ft Ft Yt -Ft
1 213 208 5 208 5 208 5
2 201 209 8 210.5 9.5 212 11
3 198 207.4 9.4 205.8 7.8 203.2 5.2
4 207 205.5 1.5 201.9 5.1 199 8
5 220 205.8 14.2 204.4 15.6 205.4 14.6
6 232 208.7 23.3 212.2 19.8 217.1 14.9
7 210 213.3 3.3 222.1 12.1 229 19
8 217 212.7 4.3 216.1 0.9 213.8 3.2
9 212 213.5 1.5 216.5 4.5 216.4 4.4
10 225 213.2 11.8 214.3 10.7 212.9 12.1
11 215.6 219.6
Total 82.3 91.0 97.4
MAD 8.23 9.10 9.74

12/8/21 NDIM 14
1.3 Trend projections
• Ex:- The sales of a company,in millions of rupees,are
given below:
Year 2001 2002 2003 2004 2005 2006 2007 2008

Sale 82 80 90 92 83 94 99 92

• Using the principle of least squares,fit a straight line

12/8/21 NDIM 15
 Year X Sales,Y XY X2 Yt Y-Yt (Y-Yt)2

2001 0 82 0 0 82 0 0

2002 1 80 80 1 84 -4 16

2003 2 90 180 4 86 4 16

2004 3 92 276 9 88 4 16

2005 4 83 332 16 90 -7 49

2006 5 94 470 25 92 2 4

2007 6 99 594 36 94 5 25

2008 7 92 644 49 96 -4 16

Total 28 712 2576 140 0 142

12/8/21 NDIM 16
 
• X =X /n=28/8=3.5

• Y = Y /n=712/8=89
 
• B= (XY)-nXY / X2 –n X 2
=2576-8*3.5*89/140-8(3.5)2
=2
 
Y X
a= -b =89-2*3=82
Accordingly the trend equation will be Yt =82+2X
with origin in 2001
12/8/21 NDIM 17
120

100

80

60 trend line
actual values

40

20

0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

12/8/21 NDIM 18
• Forecasting
• Y 2009 =82+2*8=Rs 98 million
• Y 2010 =82+2*9=Rs 100 million

12/8/21 NDIM 19
2.Casual model of forecasting
• In this model the variable to be forecasted,called the
dependent variable,is related to some
variables,known as independent variable.This
forecast is done with the help of regression analysis.
• There are two types of regression analysis:
2.1 simple regression analysis
2.2 multiple regression analysis

12/8/21 NDIM 20
2.1 Simple regression analysis
• Ex:-A car manufacturer has recently held 3-day road
side exhibits on the introduction of a new model of
its deluxe cars.The number of sales personnel
employed at each of a sample of 10 exhibitions and
the number of cars booked at each one are given as
follows:
No of 5 8 6 8 9 3 5 4 6 6
Sales men
No. of 132 160 148 156 168 102 142 98 152 142
cars
booked

12/8/21 NDIM 21
Question contd…
• Using the data,regress the number of cars booked on
the number of salesman,and obtain the regression
equation.Estimate the number of cars booked if 10
salesman are employed on an exhibition.

12/8/21 NDIM 22
Solution..
X Y XY X2 Y2
5 132 660 25 17424
8 160 1280 64 25600
6 148 888 36 21904
8 156 1248 64 24336
9 168 1512 81 28224
3 102 306 9 10404
5 142 392 25 20164
4 98 392 16 9604
6 152 912 36 23104
6 142 852 36 20164
Total 60 1400 8760 392 200298

12/8/21 NDIM 23
 
• X = X /n=60/10=6

• Y = Y /n=1400/10=140
  
• b= XY -n X Y   X 2 –n X2
=8760-10*6*140/392-10*62
=360/32
=11.25
 
a= Y -b X = 140-11.25*6=72.5

12/8/21 NDIM 24
• Thus the regression equation on Y is Yc =72.5+11.25X
• Forecast :for X=10,we have Y=72.5+11.25*10=185
• Thus ,when 10 sales persons are employed,a total of
185 cars can be expected to be booked.

12/8/21 NDIM 25
2.2 Multiple Regression Analysis
• Ex:-Given the following data:
Performance evaluation(Y): 28 33 21 40 38 46
Aptitude Test score (X1): 74 87 69 69 81 97
Prior Experience (X2): 5 11 04 09 07 10
(i) Develop the estimating equation best describing
these data.
(ii) If an employee scored 83 on the aptitude test and
had a prior experience of 7,what performance
evaluation would be expected?

12/8/21 NDIM 26
 X X1 X2 X1Y X2Y X1X2 X 12 X22

28 74 5 2072 140 370 5476 25

33 87 11 2871 363 957 7569 121

21 69 4 1449 84 276 4761 16

40 69 9 2760 360 621 4761 81

38 81 7 3078 266 567 6561 49

46 97 10 4462 460 970 9409 100

206 477 46 16692 1673 3761 38537 392

12/8/21 NDIM 27
• Accordingly, the three equations are:
206=6a+477b1+46b2
16692=477a+38537b1+3761b2
1673=46a+3761b1+392b2
Solving the three equations we get a=4.364, b1 =0.1978
and b2 =1.858
The regression equation is
Yc = 4.364+0.1978X1+1.858X2

12/8/21 NDIM 28
 Thank you

12/8/21 NDIM 29