reparation

Study Customized Readings for QNT5040 Business Modeling Chapter 9

You are to create a 12 month forecast, from September 2013 through August 2014 for Tech Tierra that uses the following forecasting techniques:

The one variable summary in StatTools.

1. The runs test to determine if the data is random or not.
2. The annual box and whisker plot.
3. The moving average with a span of 3.
4. The simple exponential smoothing forecast.
5. The Holtz linear method for trends.
6. The Winters method for trends and seasonality.

To decide which of these has the best forecast probability you will concentrate on the following results as provided by the software:

1. The root mean square error (RMSE)
2. The mean absolute percentage of error (MAPE)

Important things you need to know

1. The company’s fiscal year goes from February of one year to January of the next year.

Your Tasks

Carry out the various statistical tests as indicated above.

1. Analyze the resulting data, and determine which of the forecasting techniques provides the best forecast, and why.
2. Report on each of these forecasting techniques in your report, explaining what you did, what you found out, and how you decided which forecast is the best.
3. Forecast the next twelve months of sales for Tech Tierra.
4. Submit your individual report on time.

Write a management report (using the required format) suggesting and justifying an appropriate decision regarding the sales forecast.  In your report, analyze the patterns of the monthly time series; discuss the properties of each forecasting model and their relevance to predicting the series; select the best forecasting model for the series; make your recommendations supported by arguments which are further supported by references to model results and tables or figures in your report.  Independent outside research is encouraged to provide relevant background information and/ or to provide more support for your arguments.