Forecasting exam by Op8sxmt


									                            Forecasting exam

1. If Professor Geurts transposed December 2003 and 2004 for the Hawaii
   data, why is this less of a problem in an exponential smoothing forecasting
   of December 2005 than if it were the Novembers that had been
   transposed. Remember, we were trying to forecast December 2005 with
   data including November 2005?

   If we wanted a mistaken value to have less impact on a forecast what
   should we do with the smoothing constant alpha (), use a high value or a
   low value? What is the range of values for alpha?

2. Why should the average forecasting error be zero over time for maximum
   accuracy? If the average error were + 3% for the last six periods, what
   should you do to increase the next periods accuracy. Explain

3. Explain seasonal factors and their impact in forecasting? Include a
   discussion of what would happen in forecasting visitors to Hawaii if
   seasonality were ignored.

4. Discuss the GIGO (garbage in garbage out) concept in forecasting time
   series models. What are some of the reasons that the Hawaii visitor data
   might be wrong? Hint (preliminary data vs. revised data is one reason)

5. What are the components (patterns) in a time series? Think of the
   components forecasting method in NCSS. An additional component is
   atypical events like 9/11. How does such an event impact forecasts and
   what can be done to lessen the impact?

6. Explain the benefits of response models and why it is difficult to develop

7. How many ways do you know for forecasting new product sales? Explain
   each and the data needed.

8. Explain how to determine forecasting accuracy?

9. Write your own question and answer?

10. Give an explanation of Bayesian forecasting. Also give and explain an
Send your answers to and include your name

To top