There are two series of data: (1) The number of residential units authorized by building permits in San Diego County, and (2) the seasonally adjusted unemployment rate in San Diego County. The data are monthly observations for the period 2010 - 2014. You should enter the values for (X) and (Y) in the table below.
1. For the building permit data, estimate both the linear trend model and the constant percentage growth rate model. Which is the better model according to the appropriate statistical measures?
2. Re-estimate the linear trend model using dummy variables to correct for monthly seasonality. Indicate the months for which there is significant seasonality and the direction of that seasonality.
3. For the unemployment rate data, smooth the data with a six-month moving average and a 12-month moving average. Based on these, what is the forecast for January 2015? Which forecast is better according to the root mean square error criterion?
4. Smooth the unemployment data using exponential smoothing, first with w = 0.30 and then with w = 0.50. Based on these, what is the forecast for January 2015? Which forecast is better according to the root mean square error criterion?