BUSINESS LOGISTICS SYSTEMS ANALYSIS

1

BUSINESS LOGISTICS SYSTEMS ANALYSIS
Lecture 22FORECASTING METHODS
H. E. METZNER, Ph.D. Department of Engineering
Management University of Missouri-Rolla

2
FORECASTING

• Long Term versus Short Term
• Operational versus Strategic
• Event Impact

3
FORECASTING METHODS

• Qualitative
• Historical Projection
• Causal

4
POPULAR FORECASTING TECHNIQUES

• Delphi
• Market research
• Panel consensus
• Sales force estimates
• Visionary forecast
• Historical analogy
• Moving average
• Exponential smoothing
• Box-Jenkins
• Time-series decomposition
• Trend projections

5
FORECASTING TECHNIQUES (contd.)

• Focus forecasting
• Spectral analysis
• Regression model
• Econometric model
• Intention-to-buy and anticipation surveys
• Input-output model
• Economic input-output model
• Leading indicators
• Adaptive filtering
• Dynamic simulation

6
TECHNIQUES FOR LOGISTICIANS

• Exponential Smoothing
• Classic Time-Series Decomposition
• Multiple-Regression Analysis

7
EXPONENTIAL SMOOTHING

• Correcting for trends
• Correcting for trends and seasonality
• Forecast error
• Monitoring forecast error

8
QUALITATIVE FORECASTING

• Visionary
• Panel consensus
• Delphi panel
• Market research
• Analogies

9
HISTORICAL PROJECTION/TIME SERIES METHODS

• Moving average
• Exponential smoothing
• Box-Jenkins
• X-11
• Spectral analysis
• Naïve forecast
• Focus forecasting

10
CAUSAL FORECASTING

• Multiple regression
• Econometric
• Input-Output
• Leading indicators
• Diffusion index

11
(No Transcript)
12
Exponential Smoothing
Where current time period
exponential smoothing constant demand at
period t forecast for period t
forecast for the period following period t
13
Exponential Smoothing Correcting for Trend
Where the additional symbols not previously
defined are trend-corrected forecast for
period t1 initial forecast for period t
trend for period t trend smoothing
constant
14
Exponential Smoothing Correcting for Trend and
Seasonality
Where symbols not previously defined are
trend and seasonally corrected forecast for
period t1 smoothing constant on the
seasonal index seasonal index for period
t the time period for one full season
15
Forecast Error
Where standard error of the forecast
actual demand in period t forecast
for period t number of forecast periods t
16
Classic Time-Series Decomposition
Classic time-series analysis combines each type
of sales variation in the following way F T x
S x C x R where F demand forecast (units or
) T trend level (units or ) S seasonal
index C cyclical index T residual index
17
Classic Time-Series Decomposition
Mathematical expression for a linear trend
line T a bt
average demand level, or trend t
time
the number of observations used in the
development of the trend line the
actual demand in time period t average
demand for N time periods average of t
over N time periods
18
Classic Time-Series Decomposition
The seasonal component of the model
Where seasonal index in time period t
trend value determined from T a bt
19
Classic Time-Series Decomposition
Forecast for time period t in the future
Where the forecasted demand in time
period t number of periods in the
seasonal cycle
20

BUSINESS LOGISTICS SYSTEMS ANALYSIS
Lecture 24FORECASTING METHODS
Assignment Q4

21
Business Systems Logistics Analysis
Inventory Strategy
Transport Strategy

• Forecasting
• Storage fundamentals
• Inventory decisions
• Purchasing and supply
• scheduling decisions
• Storage decisions

• Transport fundamentals
• Transport decisions

Customer Service Goals

• The product
• Logistics service
• Information systems

Location Strategy

• Location decisions
• The network planning process

Next Lecture Special Logistics Problems