Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
FORECASTING
OUTLINE
1.0 Introduction
2.0 Types of Forecasts by Time Horizon
3.0 Steps in the Forecasting Process
4.0 Qualitative Forecasting Methods
5.0 Quantitative Forecasting Methods
1.0 INTRODUCTION
DEFINITION OF FORECASTING
Timely
Reliable Accurate
Written
Realities of Forecasting
Forecasts are seldom perfect
Most forecasting methods assume that
there is some underlying stability in the
system
Both product family and aggregated
product forecasts are more accurate
than individual product forecasts
2.0 TYPES OF FORECASTS
BY TIME HORIZON
Types of Forecasts by Time Horizon
Short-range forecast
Short-
Up to 1 year; usually less than 3 months
Job scheduling, worker assignments
Medium--range forecast
Medium
3 months to 3 years
Sales & production planning, budgeting
Long--range forecast
Long
3+ years
New product planning, facility location
Short--term vs. Longer
Short Longer--term Forecasting
HDTV
Cost minimization
design changes Competitive product changes
improvements and options Over capacity in the
Short production runs Optimum capacity
industry
Increase capacity Increasing stability of
High production costs
Shift toward product process Prune line to eliminate
Limited models focused items not returning good
Long production runs
Attention to quality Enhance distribution margin
Product improvement and
cost cutting Reduce capacity
3.0 STEPS IN FORECASTING
PROCESS
Steps in the Forecasting Process
The forecast
1. Nave approach
2. Simple Moving Averages
3. Weighted Moving Time-series
Models
Average
4. Exponential smoothing
5. Trend Projections
Associative
models
1. Linear regression (Causal)
5.0 QUANTITATIVE FORECASTING
METHODS
(TIME SERIES MODEL)
What is a Time Series?
Set of evenly spaced numerical data
(weekly, monthly etc)
Obtained by observing response variable at regular
time periods
Forecast based only on past values
Assumes that factors influencing past and present
will continue influence in future
Example
Year: 1998 1999 2000 2001 2002
Sales: 78.7 63.5 89.7 93.2 92.1
1. Naive Approach
Assumes demand in next
period is the same as
demand in most recent
period
e.g., If May sales were 48,
then June sales will be 48
Sometimes cost effective
& efficient
1995 Corel Corp.
Naive Forecasts
Uh, give me a minute....
We sold 250 wheels last
week.... Now, next week
we should sell....
Ai
MAn = i=1
n
Simple Moving Average Example
Youre manager of a museum store that
sells historical replicas. You want to
forecast sales for 2003 using a 3-period
moving average.
1998 4
1999 6
2000 5
2001 3
2002 7
Moving Average Solution
Time Response Moving Moving
Yi Total Average
(n=3) (n=3)
1998 4 NA NA
1999 6 NA NA
2000 5 NA NA
2001 3 4+6+5=15 15/3 = 5
2002 7
2003 NA
Moving Average Solution
Time Response Moving Moving
Yi Total Average
(n=3) (n=3)
1998 4 NA NA
1999 6 NA NA
2000 5 NA NA
2001 3 4+6+5=15 15/3 = 5
2002 7 6+5+3=14 14/3=4 2/3
2003 NA
Moving Average Solution
Time Response Moving Moving
Yi Total Average
(n=3) (n=3)
1998 4 NA NA
1999 6 NA NA
2000 5 NA NA
2001 3 4+6+5=15 15/3=5.0
2002 7 6+5+3=14 14/3=4.7
2003 NA 5+3+7=15 15/3=5.0
3. Weighted Moving Average Method
Period 1 2 3 4 5
Demand 42 40 43 40 41
Solution:
a. F6 = [(0.4(41)
[(0.4(41) + 0.3(40) + 0.2(43) + 0.1(40
0.1(40)]
)] / 1 = 41.0
b. F7 = [0.4(39)
[0.4(39) + 0.3(41) + 0.2(40) + 0.1(43
0.1(43)]
)] / 1 = 40.2
Actual Demand, Moving Average, Weighted
Moving Average
35 Weighted moving average
30
Actual sales
25
Sales Demand
20
15
10
Moving average
5
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Disadvantages of
Techniques for Averaging Methods
Increasing n makes forecast less
sensitive to real changes in the data
Cannot pickup trends very well.
Because they are averages, they
will always stay within past levels
and will not predict changes to
either higher or lower levels.
Require much historical
data
3. Exponential Smoothing
Ft = Ft-1 + (At-1 - Ft-1)
100
50
0
1 2 3 4 5 6 7 8 9
Quarter
4. TREND PROJECTIONS
A time-
time-series forecasting method that fits a
trend line to a series of historical data
points and then projects the line into the
future for forecasts.
Several mathematical trend equations can
be developed (e.g. linear, exponential,
quadratic)
We will focus on linear trend only
Linear Trend Projection
Deviation Deviation
Deviation
Deviation Point on
regression
Deviation line
Deviation
Y = a + bx
Time
Linear Trend Projection Model
Y$i = a + bX i
Y b>0
a
b<0
a
Time, X
Equations
Equation: Yi = a + bx i
n
x i y i nx y
Slope: b = i =1n
x i2 nx 2
i =1
Y-Intercept: a = y bx
Computation Table
2 2
Xi Yi Xi Yi X iY i
2 2
X1 Y1 X1 Y1 X 1Y 1
2 2
X2 Y2 X2 Y2 X 2Y 2
: : : : :
2 2
Xn Yn Xn Yn X nY n
2 2
Xi Yi Xi Yi X iY i
Using a Trend Line
The demand for
Year Demand
electrical power at
1997 74 N.Y.Edison over the
1998 79 years 1997 2003 is
1999 80 given at the left. Find
the overall trend.
2000 90
2001 105
2002 142
2003 122
Finding a Trend Line
Year Time Power x2 xy
Period Demand
(x) (y)
1997 1 74 1 74
1998 2 79 4 158
1999 3 80 9 240
2000 4 90 16 360
2001 5 105 25 525
2002 6 142 36 852
2003 7 122 49 854
x=28 y=692 x2=140 xy=3,063
The Trend Line Equation
x 28 y 692
x= = =4 y= = = 98.86
n 7 n 7
160
Trend line
150
Y = 56.70 + 10.54x
140
130
120
110
100
90
80
70
60
1997 1998 1999 2000 2001 2002 2003 2004 2005
Year
Linear Trend Equation (2nd Technique)
y = a + bx
0 1 2 3 4 5 x
y = Forecast for period x
x = Specified number of time periods
a = Value of Ft at t = 0
b = Slope of the line
Calculating a and b
n (xy) - x y
b =
2 2
n x - ( x)
y - b x
a =
n
Linear Trend Equation Example
t y
2
Week t Sales ty
1 1 150 150
2 4 157 314
3 9 162 486
4 16 166 664
5 25 177 885
2
t = 15 t = 55 y = 812 ty = 2499
2
( t) = 225
Linear Trend Calculation
812 - 6.3(15)
a = = 143.5
5
y = 143.5 + 6.3x
5.0 QUANTITATIVE FORECASTING
METHODS
(ASSOCIATIVE / CAUSAL MODEL)
ASSOCIATIVE MODEL
Unlike time-
time-series method, associative
model usually consider several variables
that are related to the quantity being
predicted
E.g.: the sale of IBM PCs might be related
to IBMs advertising budget, the
companys prices, competitors prices etc.
PC sales = dependent variable;
other variables = independent variables
Main technique: Linear Regression
Linear Regression Model
Shows linear relationship between
dependent & explanatory variables
Example: Sales & advertising (not
(not time)
Y-intercept Slope
Y^ i = a + b X i
Dependent Independent (explanatory)
(response) variable variable
Example:
Expenditure on advertising is expected to cause
an increase in sales volume, or
Sales Volume = A + B x advertising expenditure
Where :
A = Sales volume without advertising
B = The amount by which one unit of advertising
expenditure can increase sales volume.
Linear Regression Model
Y Y i = a + b X i +Error
Error
Regression line
Y^i = a + b X i
X
Observed value
Equations
Equation: Yi = a + bx i
n
x i y i nx y
Slope: b = i =1n
x i2 nx 2
i =1
Y-Intercept: a = y bx
or
n (xy) - x y
b =
2 2
n x - ( x)
y - b x
a =
n
Scatter Diagram
3
th o u s a n d s )
0
0 1 2 3 4 5 6 7 8
Area Payroll (in $ hundreds of millions)
EXAMPLE
Data shows sales of 20 TV and unemployment. Derive
predictive equation for sales based on unemployment
levels.
Period Unit Sold Unemployment (%)
1 20 7.2
2 41 4.0
3 17 7.3
4 35 5.5
5 25 6.8
6 31 6.0
7 38 5.4
8 50 3.6
9 15 8.4
10 19 7.0
11 14 9.0
Solution:
a ) Plot a graph data
60
50
Unit Sold
40
Unit Sold vs Level of
30
Unemployment
20
10
0
0.0 2.0 4.0 6.0 8.0 10.0
Level of Unemployment (%)
b) Calculate x, y, xy, x2, y2
b = 11 (1750.8) 70.2(305)
11 (476.4) 70.2 (70.2)
= -2152.2 / 312.36 = -6.89
y = 71.85 - 6.89x
Additional exercises