A call center operates from 8:00am to 5:00pm, Monday through Friday. The actual number of calls (demand) for each hour of the day and each day of the week has been recorded for the past 16 weeks. [Note: there are 9 hours per day, 5 days per week, and 16 weeks of data. Therefore, there are 9 5 16 = 720 demand values in the time series]. The data are available in the Forecasting worksheet of the Assignment 1 workbook. Your goal is to build a time series forecasting model for predicting call volume. You should only use the first 14 weeks (630 data points) of the time series for building the model. The last two weeks (90 data points) should be used to evaluate the performance of the predictive model. [Think of the first 14 weeks as a training sample for building the model and the last two weeks as a validation sample for evaluating the model].

When building and evaluating your forecasting model, you should report mean absolute deviation (MAD), mean square error (MSE), and mean absolute percentage (MAPE) error as measures of fit. You should also establish control limits to evaluate your predictions for the validation sample. (1) Briefly describe the process that you used to develop your time series forecasting model. (2) Carefully describe the time series forecasting model that you selected. For example, a simple exponential smoothing model using alpha = 0.27 was selected based on minimum MSE, or a classical decomposition model using an intercept of 200, slope of 1, and seasonal weights of ... was selected based on minimum MAPE. (3) Report the MAD, MSE, and MAPE for the training sample for your selected model. (4) Report the MAD, MSE, and MAPE for the validation sample for your selected model. (5) What percentage of the validation sample cases fall within your control limits? (6) Provide a qualitative assessment of the predictive accuracy of your model?