Professional Documents
Culture Documents
for business
CHAPTER 5
FORECASTING
Chapter Outline
5.1 Introduction
5.2 Types of Forecasts
5.3 Scatter Diagrams
5.4 Measures of Forecast Accuracy
5.5 Time-Series Forecasting Models
5.6 Monitoring and Controlling Forecasts
5.7 Using the Computer to Forecast
Introduction
Why forecasting
are important?
Introduction 8 steps to forecasting:
Step 1: Determine the Step 2: Select the Step 3: Determine the Step 4: Select the
use of the forecast— forecasted items or time horizon of the forecasting model or
objective quantities forecast models
Step 8: Implement Step 7: Make Step 6: Validate the Step 5: Gather the
results the forecast forecasting model data needed to
make the forecast
Forecasting Models
Forecasting
Techniques
300
6 250 350 150 250
200
7 250 360 160 150
100
8 250 370 190 50
0
9 250 380 200 0 2 4 6 8 10 12
Forecasts
Demands
MAD
forecast error
160
17.8
n 9
Measures of Forecast Accuracy
We compare forecasted values with actual values to see how well one model
works or to compare models
Forecast error = Actual value – Forecast value
One measure of accuracy is the mean absolute deviation (MAD)
MAD
MAD
forecast error
n
There are other popular measures of forecast accuracy
The mean squared error
MSE
( error) 2
n
The mean absolute percent error
error
actual
MAPE 100%
n
bias
error
n
TIME SERIES FORECASTING MODELS
Trend
Component
Seasonal Peaks
Actual
Demand
Line
Average Demand
over 4 Years
| | | |
1. Moving Averages
used when demand is relatively steady over time
The next forecast is the average of the most recent n data values from
the time series
Tend to smooth out short-term irregularities in the data series
Ft 1
( Weight in period i )( Actual value in period)
( Weights )
Mathematically
w1Yt w2Yt 1 ... w nYt n1
Ft 1
w1 w2 ... w n
ere
wi= weight for the ith observation
TIME SERIES FORECASTING MODELS
6
Sum of the weights
TIME SERIES FORECASTING MODELS
2. Weighted Moving Averages
THREE-MONTH WEIGHTED
MONTH ACTUAL SHED SALES MOVING AVERAGE
January 10 Forecasted
February 12 values
March 13
April 16 [(3 X 13) + (2 X 12) + (10)]/6 = 12.17
May 19 [(3 X 16) + (2 X 13) + (12)]/6 = 14.33
3. Exponential Smoothing
Exponential smoothing is easy to use and
requires little record keeping of data
It is a type of moving average
3. Exponential Smoothing
Exponential smoothing forecast for two values of
ACTUAL
TONNAGE FORECAST FORECAST
QUARTER UNLOADED USING =0.10 USING =0.50
1 180 175 (Any reasonable starting figure) 175
2 168 175.5 = 175.00 + 0.10(180 – 175) 177.5
3 159 174.75 = 175.50 + 0.10(168 – 175.50) 172.75
4 175 173.18 = 174.75 + 0.10(159 – 174.75) 165.88
5 190 173.36 = 173.18 + 0.10(175 – 173.18) 170.44
6 205 175.02 = 173.36 + 0.10(190 – 173.36) 180.22
7 180 178.02 = 175.02 + 0.10(205 – 175.02) 192.61
8 182 178.22 = 178.02 + 0.10(180 – 178.02) 186.30
9 ? 178.60 = 178.22 + 0.10(182 – 178.22) 184.15
8 182
Best178.22
choice 3.78
186.30
4.3..
Sum of absolute deviations
develop trial forecasts82.45
with different values of 98.63
select the that results in the lowest MAD
andΣ|deviations|
MAD = = 10.31 MAD = 12.33
TIME SERIES FORECASTING MODELS
4. Exponential Smoothing with Trend Adjustment
Forecast including trend (FITt+1) = new forecast (Ft+1) + trend correction (Tt+1)
FITt 1 Ft 1 Tt 1
Ft 1 Ft Yt Ft Tt 1 (1 )Tt ( Ft 1 Ft )
Unadjusted Adjusted
Period Demand Forecast Ft Trend Tt Forecast FITt
1 54 50 0 -
2 57 50.8 0.56 51.36
3 44 - - -
2017 90
2018 105
2019 142
2020 122
2021 ???
Midwestern Manufacturing
Company Example
TIME SERIES FORECASTING MODELS
Yˆ b0 b1 X
= predicted
Ŷ value
b0= intercept
b1= slope of the line
X= time period (i.e., X = 1, 2, 3, …, n)
TIME SERIES FORECASTING MODELS
Dist5 * (X X ) 2
Dist6
* Dist3 * b0 Y b1 X
Dist4
X average (mean) of X values
Dist1 * Dist2
* X
n
* Y
Y average (mean) of Y values
n
Time
TIME SERIES FORECASTING MODELS
Notice code
instead of
actual years
1,200 1,200
Jan. 0.957 96 July 1.117 112
12 12
1,200 1,200
Feb. 0.851 85 Aug. 1.064 106
12 12
1,200 1,200
Mar. 0.904 90 Sept. 0.957 96
12 12
1,200 1,200
Apr. 1.064 106 Oct. 0.851 85
12 12
1,200 1,200
May 1.309 131 Nov. 0.851 85
12 12
1,200 1,200
June 1.223 122 Dec. 0.851 85
12 12
TIME SERIES FORECASTING MODELS
Seasonal
Definite trend pattern
Turner Industries Example
TIME SERIES FORECASTING MODELS
7. Seasonal Variations with Trend
Step 1 Step 2
YEAR QUARTER SALES CMA SEASONAL RATIO = Observation/CMA
for that observation
1 1 108
2 125
3 150 132.000 1.136
4 141 134.125 1.051 The sum of these
2 1 116 136.375 0.851 indices should be the
number of seasons (4)
2 134 138.875 0.965 since an average
3 159 141.125 1.127 season should have an
index of 1. In this
4 152 143.000 1.063 example, the sum is 4.
3 1 123 145.125 0.848 If a sum were not 4, an
adjustment would be
2 142 147.875 0.960 made by multiply each
index by 4 and divide
3 168
this by the sum of the
4 165 indices.
Step 3: Calculate seasonal indices
Index for quarter 1 = I1 = (0.851 + 0.848)/2 = 0.85
There are two seasonal
ratios for each quarter so Index for quarter 2 = I2 = (0.965 + 0.960)/2 = 0.96
these are averaged to get Index for quarter 3 = I3 = (1.136 + 1.127)/2 = 1.13
the seasonal index
Index for quarter 4 = I4 = (1.051 + 1.063)/2 = 1.06
TIME SERIES FORECASTING MODELS
7. Seasonal Variations with Trend
Suppose the FOURTH year’s annual demand for 640
UNITS (one year), what are the QUARTERLY demand?
Q1
Q2
Q3
Q4
TIME SERIES FORECASTING MODELS
8. The Decomposition Method of Forecasting
YEAR QUARTER
SALES SEASONAL DESEASONALIZED
($1,000,000s) INDEX SALES ($1,000,000s)
1 1 108 0.85 127.059
2 125 0.96 130.208
3 150 1.13 132.743
4 141 1.06 133.019
2 1 116 0.85 136.471
2 134 0.96 139.583
3 159 1.13 140.708
4 152 1.06 143.396
3 1 123 0.85 144.706
2 142 0.96 147.917
3 168 1.13 148.673
4 165 1.06 155.660
TIME SERIES FORECASTING MODELS
8. The Decomposition Method of Forecasting
Step 3: Find the equation of a trend line using the deseasonalized data
Find a trend line using the deseasonalized data
b1 = 2.34 b0 = 124.78
Step 4: Develop a forecast using this trend a multiply the forecast by
the appropriate seasonal index
Ŷ = 124.78 + 2.34X
= 124.78 + 2.34(13)
= 155.2 (forecast before adjustment for
seasonality)
Step 5: Multiply the trend line forecast by the appropriate seasonal
index
Ŷ x I1 = 155.2 x 0.85 = 131.92
TIME SERIES FORECASTING MODELS
9. Regression with Trend and Seasonal
Components
Multiple regression can be used to forecast both
trend and seasonal components in a time series
One independent variable is time
Dummy independent variables are used to represent
the seasons
The model is an additive decomposition model
Yˆ a b1 X 1 b2 X 2 b3 X 3 b4 X 4
where X1 = time periods
X2 = 1 if quarter 2, 0
otherwise
X3 = 1 if quarter 3, 0
otherwise
X4 = 1 if quarter 4, 0
otherwise
TIME SERIES FORECASTING MODELS
9. Regression with Trend and Seasonal
Components
TIME SERIES FORECASTING MODELS
9. Regression with Trend and Seasonal
Components
TIME SERIES FORECASTING MODELS
9. Regression with Trend and Seasonal
Components Intercept
X Variable 1
104.1041667
2.3125
where
MAD
forecast error
n
Kimball’s Bakery Example
Tracking signal for quarterly sales of croissants
MAD
forecast error 85
14.2
n 6
RSFE 35
Tracking signal 2.5MADs
MAD 14.2
Monitoring and Controlling Forecasts
Signal Tripped
Upper Control Limit Tracking Signal
+
Acceptable
0 MADs Range
–
Lower Control Limit
Time
Exponetial Smoothing:
Ft+1= new forecast (for time period t + 1)
Ft 1 Ft where
(Yt Ft ) Ft= pervious forecast (for time period t)
= smoothing constant (0 ≤ ≤ 1)
Yt= pervious period’s actual demand
Summary
Exponetial Smoothing with trend adjustment:
where
FITt 1 Ft 1 Tt 1 Tt+1 = smoothed trend for period t + 1
Tt = smoothed trend for preceding
Ft 1 Ft Yt Ft period
= trend smooth constant that we
Tt 1 (1 )Tt ( Ft 1 Ft ) select
Ft+1 = simple exponential smoothed
forecast for period t + 1
Trend Projection: Ft = forecast for pervious period
Yˆ b0 bwhere
1X
Ŷ
= predicted value
X = time period (i.e., X = 1, 2, 3, …, n)
b1
( X X )(Y Y )
X average (mean) of X values
(X X ) 2
X
n
b0 Y b1 X Y
Y average (mean) of Y values
n
Summary
Four steps in computing seasonal indices
1. Compute the CMA for each observation (where possible)
2. Compute the seasonal ratio = Observation/CMA for that observation
3. Average seasonal ratios to get seasonal indices
4. If seasonal indices do not add to the number of seasons, multiply each
index by (Number of seasons)/(Sum of indices)
Five steps to decomposition
1. Compute seasonal indices using CMAs
2. Deseasonalize the data by dividing each number by its seasonal index
3. Find the equation of a trend line using the deseasonalized data
4. Forecast for future periods using the trend line
5. Multiply the trend line forecast by the appropriate seasonal index
Additive decomposition model
X1 = time period
Yˆ a b1 X 1 b2 X 2 b3 X 3 b4 X 4 X2, X3, X4 = 1 if quarter 2,3, 4;
0 otherwise
Homework