Lecture 2 - University of Alberta

Download Report

Transcript Lecture 2 - University of Alberta

The Operations Management Club organizes industry mixers,
seminars, technical workshops, and conferences for students with
an interest in Operations Management and Management Science.
If you are interested in joining the OM Club, or are considering a
major in Operations Management and have any questions about
the degree, we would like to hear from you.
For more information on the club, membership, and events, visit
http://studentweb.bus.ualberta.ca/om/
or email [email protected]
Meeting: Tuesday, January 16 at 5:00 PM, Bus 4-10
Announcements
• HW 1 due Wednesday, 11:59 PM
• OM Club Excel workshops
– Jan 20, 11 AM – 1 PM
– Free
– Watch for a sign up link on the course page
• Don’t have course pack yet?
– Get one Friday in Lab
MGTSC 352
Lecture 2: Forecasting
Why forecast?
Types of forecasts
“Simple” time series forecasting methods
Including SES = Simple Exponential Smoothing
Performance measures
Plant Site Selection
• Alberta Manufacturer
• Has one old plant, in Calgary
• Planning to build new plant, but where?
– Edmonton or Calgary?
Recent Demand Figures
Calgary
Edmonton
Fort McMurray
Red Deer
2001
2002
2003
2004
2005
What Would you Do?
Perspectives on Forecasting
• Forecasting is difficult, especially if it's about
the future!
Niels Bohr
• Rule #0: Every forecast is wrong!
– Provide a range
More sarcastic quotes about forecasting:
http://www.met.rdg.ac.uk/cag/forecasting/quotes.html
What is the Driver Doing?
Forecasting
• Technological forecasts
– New product, product life cycle
(Ipod, Blackberry)
– Moore’s Law
– Gates’ Law
• Economic forecasts
– Macro level (unemployment, inflation, markets, etc.)
• Demand forecasts
– Focus in MGTSC 352
Moore's Law: Computing power
doubles about every two years.
1,000,000,000
100,000,000
Transistors
10,000,000
1,000,000
Gates’
Law: “The speed of software
halves every 18 months.”
100,000
10,000
1,000
1965
1975
1985
Year
1995
2005
Data from ftp://download.intel.com/museum/Moores_Law/Printed_Materials/Moores_Law_Backgrounder.pdf
Economic Forecasts
An economist is an expert who will
know tomorrow why the things he
predicted yesterday didn't happen
today.
Evan Esar
Why do economists make forecasts?
“We forecast because people with money ask us to.”
Kenneth Galbraith
Forecasting – Quantitative
• Time series analysis: uses only past
records of demand to forecast future demand
– moving averages
– exponential smoothing
– ARIMA
• Causal methods: uses explanatory variables
(timing of advertising campaigns, price changes)
– multiple regression
– econometric models
Active learning
• Groups of two
• Recorder: person that is born closest to
Telus 150.
• Task: think of three quantities that you’d like
to forecast
• 1 minute
Choosing a Forecasting Method
START
Is forecast
important?
Yes
No
Are accurate historical
data available?
No
Yes
Is forecast
important?
Yes
No
Is there at least 1-2
Flip a coin;
months before
use your intuition;
forecast is needed? No look at your horoscope;
consult an economist
Yes
Select appropriate
qualitative method
Are you willing to
pay for greater
accuracy?
Yes
No
Use a causal Use time-series
method
method
END
Simple models
• Notation
– Dt = Actual demand in time period t
– Ft = Forecast for period t
– Et = Dt - Ft = Forecast error for period t
• Problem: Forecast the TSX index
4 simple models
Excel
(Simple) Exponential Smoothing
• Generalization of the WMA method
• Uses a single parameter for weights
0  LS  1
• Three steps
– Initialization ... F2 = D1
– Calibration ... Ft+1 = LS Dt + (1 - LS) Ft
– Prediction ... same formula
Note the formula is a weighted average of Demand and Forecast from last period
Excel
SES weights
• Decrease “exponentially” as data age
• Most recent data gets a weight of LS
• Ft+1 = [LS Dt ] + [(1 - LS) Ft ]
Rearrange...
• Ft+1 = Ft + LS (Dt - Ft)
= Ft + LS Et
• A learning model
How do we choose LS
• Active learning (1 min.):
– High LS (≈ 1) results in ....
– Low LS (≈ 0) results in ....
• Suggested range for LS: (0.01,0.3)
• Performance measures (formulas in course pack, pg. 21)
–
–
–
–
–
BIAS
MAD
SE
MSE
MAPE
Excel
Famously Incorrect Forecasts
• “I think there is a world market for maybe five
computers.”
Thomas Watson, chairman of IBM, 1943
• “There is no reason anyone would want a
computer in their home.”
Ken Olson, president, chairman and founder of Digital Equipment
Corp., 1977
• “The concept is interesting and well-formed, but
in order to earn better than a 'C,' the idea must
be feasible.”
A Yale University management professor in response to Fred
Smith's paper proposing reliable overnight delivery service. (Smith
went on to found Federal Express Corp.)