D.
Anthony Chevers
SBCO 6240 - Production and Operations Management
Lecture #4 – Forecasting
Definition & purpose
Forecasting models
Simple moving average
Weighted moving average
Exponential smoothing
Trend projections
Regression analysis
Forecast errors
Exercises
Lecture 6 – Forecasting | 2
My interest is in the future
because I am going to spend
the rest of my life there.
—Charles F. Kettering
Lecture 6 – Forecasting | 3
Some Reasons Why
Forecasting is Essential
in OM
New Facility Planning – It can take 5 years to design
and build a new factory or design and implement a new
production process.
Production Planning – Demand for products vary from
month to month and it can take several months to
change the capacities of production processes.
Workforce Scheduling – Demand for services (and the
necessary staffing) can vary from hour to hour and
employees weekly work schedules must be developed
in advance.
Lecture 6 – Forecasting | 4
Forecasting - Defined
Forecasting is the prediction/estimation of future
activities. Types of forecast – Economic, Technological and
Demand
• The first step in planning is forecasting, or estimating the future
demand for products and services and the resources necessary
to produce these outputs. E.g. Rising Star finals (2005); Craft Village –New Kgn for World Cup Cricket
2007 -closure & Diana Ross (Jazz Festival, Jan 2008); Girl you have Intimate potential – Gregory Isaacs (2008)
• Operations managers need long range forecasts to make
strategic decisions about products, processes and facilities.
• Operations managers need short range forecasts to assist
them in making decisions about production issues that span
only the next few weeks.
Lecture 6 – Forecasting |
Factors to Consider
Costs associated with a model
Required accuracy
Relevance of past data & availability
Forecast horizon
Pattern of data
Time needed for analysis
Lecture 6 – Forecasting |
6
Introduction
Eight steps to forecasting:
• Determine the use of the forecast
• Select the items or quantities to be forecasted
• Determine the time horizon of the forecast
• Select the forecasting model or models
• Gather the data needed to make the forecast
• Validate the forecasting model
• Make the forecast
• Implement the results
Lecture 6 – Forecasting |
Forecasting Models
Forecasting
Techniques
Qualitative Time Series Causal Methods
Models Methods
Delphi Moving Regression
Methods Average Analysis
Jury of Executive Exponential Multiple
Opinion Smoothing Regression
Sales Force Trend Projections
Composite
Consumer
Market Survey
Lecture 6 – Forecasting |
Lecture 6 – Forecasting |
Method
Subjective Models
Lecture 6 – Forecasting |
Some Time Series Models
Simple Moving Average
K-Period MA =
∑(Actual Value in previous k
periods) / k
Weighted moving average
k
K-Period WMA = k
i=1
i=1
∑(weight for period i) (actual value in period i)
∑(weights)
Lecture 6 – Forecasting | 11
Exponential Smoothing
Exponential Smoothing method: A sophisticated weighted moving
average method that calculates the average of a time series by
giving recent demands more weight than earlier demands.
Exponential smoothing is the most frequent used formal forecasting
method because of its simplicity and the small amount of date
needed to support it.
(Demand this period) (Forecast calculated last period)
SES: Ft+1 = α + (1-α)
= α Dt + (1-α)Ft
Or an equivalent equation: Ft+1 = Ft +α(Dt-Ft)
Where α is a smoothing parameter (0 ≤ α ≤ 1).
Lecture 6 – Forecasting | 12
Equations - Simplified
1 SM
Lecture 6 – Forecasting |
Calculation of 3-Month
Simple Moving Average
Month Actual Shed 3-Month Moving Average
Sales
January 10
February 12
March 13
April 16 (10+12+13)/3 = 11 2/3
May 19 (12+13+16)/3 = 13 2/3
June 23 (13+16+19)/3 = 16
July 26 (16+19+23)/3 = 19 1/3
Lecture 6 – Forecasting | 14
Limitations of Simple MA
Month Actual 3 Month MA
Sales
January 13
February 12
March 10
April 16 (13 + 12 + 10)/3 = 11.67
Simple MA does not capture trends nor seasonalities
Lecture 6 – Forecasting | 15
Calculating Weighted
Moving Averages
Weights Period
Applied
3 Last month
2 Two months ago
1 Three months ago
3 * Sales last moth +
2 * Sales two months ago +
1 * Sales three months ago
6 Sum of weights
Lecture 6 – Forecasting | 16
Calculation of 3-Month
Weighted-Moving Average
Month Actual 3 Month MA
Shed
Sales
January 10
February 12
March 13
April 16 [(3*13) + (2*12) + (1*10)] 6 = 12 1
6
May 19 [(3*16) + (2*13) + (1*12)] = 14 1
3
June 23 [(3*19) + (2*16) + (1*14)] = 17
July 26 [(3*23) + (2*19) + (1*16)] = 20 ½
Lecture 6 – Forecasting | 17
Exponential Smoothing
New forecast =
previous forecast + α (previous actual - previous)
or:
where
F t = F t −1 +α ( A t −1 − F t −1 )
F t = new forecast
F t-1 = previous forecast
α = smoothing constant
A t −1 = previous period actual
Lecture 6 – Forecasting | 18
Equation [Ft = Ft-1 + α (At-1 – Ft-1 )]
TABLE 5.4
Port of Baltimore Exponential Smoothing Forecasts for a α = 0.10 and α = 0.50
QUARTER ACTUAL ROUNDED FORECAST USING ROUNDED
TONNAGE α =0.10* FORECAST USING
UNLOADED α =0.50*
1 180 175 175
2 168 176 = 175.00 + 1.10(180 – 175) 178
3 159 175 =175.50 + 0.10(168-175.50) 173
4 175 173 =174.75 + 0.10(159-174.75) 166
5 190 173 =173.18 + 0.10(175-173.18) 170
6 205 175 =173.36 + 0.10(190-173.36) 180
7 180 178 =175.02 + 0.10(205-175.02) 193
8 182 178 =178.02 + 0.10(180-178.02) 186
9 ? 179= 178.22 + 0.10(182-178.22) 184
* Forecasts rounded to the nearest ton
Lecture 6 – Forecasting | 19
Selecting the Smoothing
Constant (α )
Select α to minimize:
Mean absolute deviation = MAD = ∑ | forecast errors |
n
2
Mean Square Error = MSE = ∑
( forecast errors )
n
1 | forecast error |
Mean Absolute Percent Error = MAPE = ∑
n actual
Bias = ∑ forecast errors
Lecture 6 – Forecasting | 20
Table – Detailed
Calculation
[84/8 = 10.50 & 100/8 = 12.50; Select: α = 0.10]
TABLE 5.5
Absolute Deviations and MADs for Port of Baltimore Example
QUARTER ACTUAL ROUNDED ABSOLUTE ROUNDED ABSOLUTE
TONNAGE FORECAST DEVIATIONS FORECAST FORECAST
UNLOADED USING FOR USING USING
α =0.10* α =0.10* α =0.50* α =0.50*
1 180 175 5 175 5
2 168 176 8 178 10
3 159 175 16 173 14
4 175 173 2 166 9
5 190 173 17 170 20
6 205 175 30 180 25
7 180 178 2 193 13
8 182 178 4 186 4
Sum of absolute deviations 84 100
●| deviations | MAD = 12.50
MAD = = 10.50
n
Lecture 6 – Forecasting | 21
Forecast Error – MSE
[n = 4 (# of periods)]
Actu
Tonn
MSE = ∑ forecast errors2
n
87.25 ?
Quarter Unloa
As a practice, find the MSE for α = 0.50 and conclude findings.
Lecture 6 – Forecasting | 22
Trend Projection
A 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 medium-to-long range
forecasts.
Look at linear (straight-line) trends only
Develop a linear trend line by a precise statistical
method, we can apply the least squares method
A straight line that minimizes the sum of the squares of
the vertical differences or deviations from the line to
each of the actual observations.
Lecture 6 – Forecasting | 23
The Least Squares Method
for Finding the Best-Fitting Straight Line [Time on x-axis]
Deviation
Values of dependent variable (values)
Actual observation (y value)
Deviation Deviation
5
6
Deviation
3
Deviation
4
Deviation
(error) Deviation
1
2
Trend Line,
y=a + bx
1 2 3 4 5 6 7
Time period
Lecture 6 – Forecasting | 24
Least Squares Line
Described in terms of its y-intercept (the height
at which it intercepts the y-axis) and its slope
(the angle of the line)
If we compute the y-intercept and slope, we
can express the line with the following
equation:
y = a + bx
where y (called “y hat”) = computed value of the variable to
be predicted (dependent variable)
a = y-axis intercept
b = slope of the regression line (or the rate of change in y for
given change in x) [Slope = y/x]
x = the independent variable (which in this case is time)
Lecture 6 – Forecasting | 25
Trend Projection –
General Regression equation
Y = a + bX
where
Y = computed value of the variable to
be predicted (dependent variable)
a = Y - axis Intercept [ a = y – bx ]
∑ XY −nXY
b=
∑ X 2 −nX 2
Lecture 6 – Forecasting | 26
Notations
ƅ = slope of the regression line
∑ = summation sign
x = known values of the independent variable
y = known values of the dependent variable
x = average of the value of the x’s
ӯ = average of the values of the y’s
n = number of data points or observations
Lecture 6 – Forecasting | 27
Exercise -
BurkeCampbellChevers
The demand for electrical power at Year Electrical Power
BurkeCampbellChevers Inc. Over the Demand
period 1999 to 2005 is shown below, in 1999 74
megawatts. Let’s forecast 2006 demand 2000 79
by fitting a straight-line trend to these
data: 2001 80
2002 90
2003 105
2004 142
2005 122
We can designate 1999 as year 1, 2000 as year 2, etc.
Lecture 6 – Forecasting | 28
Solution:
b = (∑XY – nXY) / (∑X2 – nX2)
Year Tim
a
1999
x=
=Y-bX
28/7
a=9
8.8
6-(
10.5
4x4
)
a=9
8.8
6-4
2.1
6
a= 5
6.
7
Lecture 6 – Forecasting | 29
Least Square Trend
Equation
Thus, the least square trend equation is:
y = 56.70 + 10.54 x
To project demand in 2006, we first denote the
year 2006 as x=8
Demand in 2006 = 56.70 + 10.54 (8)
141.02 or 141 megawatts
Estimate for 2007 – Year=9 (x=9)
Demand in 2006 = 56.70 + 10.54 (9)
151.56 or 152 megawatts
Lecture 6 – Forecasting | 30
Electrical Power and the
Computed Trend Line
Trend Line,
160 y= 56.70 + 10.54x
150
140
130
120
Power demand
110
100
90
80
70
60
50
1997 1998 1999 2000 2001 2002 2003 2004 2005
Year
Lecture 6 – Forecasting | 31
Regression Analysis
Causal forecasting models usually consider
several variables that are related to the
quantity being predicted
This approach is more powerful than the time-
series methods that use only the historical
values for the forecasting variable
Lecture 6 – Forecasting | 32
Regression Analysis
For example, the sales of PC’s might be related to
advertising budget, the price charged, competitors’
prices, promotional strategies, the economy,
disposable income, unemployment rates or time.
In this case, PC sales would be called the dependent
variable (Y) and the other variables would be called
independent variables (X).
We will use Y with one other variable (X)
The most common quantitative casual forecasting
model is linear-regression analysis
Lecture 6 – Forecasting | 33
Linear Regression
Analysis
A straight-line mathematical model to describe
the functional relationships between
independent and dependent variables
Can use the same mathematical model
employed in least squares method of trend
projection to perform a linear regression
analysis
The dependent variable that we want to
forecast is still y-hat but the independent
variable, x, need no longer be time.
Lecture 6 – Forecasting | 34
Using Regression Analysis to
Forecast
Question: If the Local Payroll is 6 in the 7th month, what is the Sales?
Y X
Triple A' Sales Local Payroll
($100,000's) ($100,000,000)
2.0 1
3.0 3
2.5 4
2.0 2
2.0 1
3.5 7
Y = 1.75 + 0.25X
Lecture 6 – Forecasting | 35
Solution
b = (∑XY – nXY) / (∑X2 – nX2)
Y X XY X2
2 1 2 1
3 3 9 9
2.5 4 10 16
2 2 4 4
2 1 2 1
3.5 7 24.5 49
15 18 51.5 80
Total
Average 2.50 3.00
If X is 6
Lecture 6 – Forecasting | 36
Solution - Projection
If the local Chamber of Commerce predicts
that the payroll will be $600 million next
period, we can estimate sales with the
regression equation:
• 1.75 + 0.25(6)
• 1.75 + 1.50 = 3.25
Thus sales = $325,000
Lecture 6 – Forecasting | 37
Exercise 1
Demand for heart transplant surgery at Washington General Hospital
has increased steadily in the past few years, as seen in the table.
Year Outpatient Surgeries
Performed
1 45
2 50
3 52
4 56
5 58
6 ?
_____________________________
The director of medical services predicted six years ago that the
demand in year 1 would be for 41 surgeries
(a) Use exponential smoothing, first with a smoothing constant of
0.6 and then with one of 0.9, to develop forecasts for years 2 through
6.
Lecture 6 – Forecasting | 38
Exercise 2
Tongren Construction Company renovates old homes in
Balvenie, Mandeville. Over time, the company has found that its
dollar volume of renovation work is dependent on the Balvenie
area payroll. The following table lists Tongren’s revenues and
the amount of money earned by wage earners in Balvenie
during the following six years.
Tongren’s Sales Local Payroll
($000,000) ($000,000,000)
2 1
3 3
2.5 4
2 2
2 1
3.5 7
Forecast sales if the payroll in period #7 is 8.
Lecture 6 – Forecasting | 39
Solution – Tutorial #1
Exponential smooth
3
Year Demand- Y e a r
1 YearD
45
Exponential smoot
2 501
Lecture 6 – Forecasting | 40
Solution – Tutorial #2
Y
2
3
2.5
Lecture 6 – Forecasting | 41
Normal Distribution
.00 .01 .02 .03 .04 .05 .06 .07 .08 .09
.0 .5000 .5040 .5080 .5120 .5160 .5199 .5239 .5279 .5319 .5359
.1 .5398 .5438 .5478 .5517 .5557 .5596 .5636 .5675 .5714 .5753
.2 .5793 .5832 .5871 .5910 .5948 .5987 .6026 .6064 .6103 .6141
∞
.3 .6179 .6217 .6255 .6293 .6331 .6368 .6406 .6443 .6480 .6517
.4
.5
.6
.6554
.6915
.7257
.6591
.6950
.7291
.6628
.6985
.7324
.6664
.7019
.7357
.6700
.7054
.7389
.6736
.7088
.7422
- .6772
.7123
.7454
.6808
.7157
.7486
0 z
.6844
.7190
.7517
.6879
+
.7224
.7549
∞
.7 .7580 .7611 .7642 .7673 .7704 .7734 .7764 .7794 .7828 .7852
.8 .7881 .7910 .7939 .7967 .7995 .8023 .8051 .8078 .8106 .8133
. 9 .8159 .8186 .212 .8238 .8264 .8289 .8315 .8340 .8365 .8389
1.0 .8413 .8438 .8461 .8485 .8508 .8531 .8554 .8577 .8599 .8621
1.1 .8643 .8665 .8686 .8708 .8729 .8749 .8770 .8790 .8810 .8830
1.2 .8849 .8869 .8888 .8907 .8925 .8944 .8962 .8980 .8997 .9015
1.3 .9032 .9049 .9066 .9082 .9099 .9115 .9131 .9147 .9162 .9177
1.4 .9192 .9207 .9222 .9236 .9251 .9265 .9279 .9292 .9306 .9319
1.5 .9332 .9345 .9357 .9370 .9382 .9394 .9406 .9418 .9429 .9441
1.6 .9452 .9463 .9474 .9484 .9495 .9505 .9515 .9525 .9535 .9545
1.7 .9554 .9564 .9573 .9582 .9591 .9599 .9608 .9616 .9625 .9633
1.8 .9641 .9649 .9656 .9664 .9671 .9678 .9686 .9693 .9699 .9706
1.9 .9713 .9719 .9726 .9732 .9738 .9744 .9750 .9756 .9761 .9767
2.0 .9772 .9778 .9783 .9788 .9793 .9798 .9803 .9808 .9812 .9817
2.1 .9821 .9826 .9830 .9834 .9838 .9842 .9846 .9850 .9854 .9857
2.2 .9861 .9864 .9868 .9871 .9875 .9878 .9881 .9884 .9887 .9890
2.3 .9893 .9896 .9898 .9901 .9904 .9906 .9909 .9911 .9913 .9916
2.4 .9918 .9920 .9922 .9925 .9927 .9929 .9931 .9932 .9934 .9936
2.5 .9938 .9940 .9941 .9943 .9945 .9946 .9948 .9949 .9951 .9952
2.6 .9953 .9955 .9956 .9957 .9959 .9960 .9961 .9962 .9963 .9964
2.7 .9965 .9966 .9967 .9968 .9969 .9970 .9971 .9972 .9973 .9974
2.8 .9974 .9975 .9976 .9977 .9977 .9978 .9979 .9979 .9980 .9981
2.9 .9981 .9982 .9982 .9983 .9984 .9984 .9985 .9985 .9986 .9986
3.0 .9987 .9987 .9987 .9988 .9988 .9989 .9989 .9989 .9990 .9990
3.1 .9990 .9991 .9991 .9991 .9992 .9992 .9992 .9992 .9993 .9993
3.2 .9993 .9993 .9994 .9994 .9994 .9994 .9994 .9995 .9995 .9995
3.3 .9995 .9995 .9995 .9996 .9996 .9996 .9996 .9996 .9996 .9997
3.4 .9997 .9997 .9997 .9997 .9997 .9997 .9997 .9997 .9997 .9998
Lecture 6 – Forecasting | 42
NEXT LECTURE:
Inventory Management
D. Anthony Chevers
[email protected]
DOMS, Room #28
43
