Refer to scenario in Problem 4 using the le BlueOrRed. Create a standard partition of the data with all the tracked variables and 50% of observations in the training set, 30% in the validation set, and 20% in the test set. Apply the random trees procedure to create an ensemble of classication trees using Age, HomeOwner, Female, Married, HouseholdSize, Income, Education, and Church as input variables and Undecided as the output variable. In Step 2 of XLMiner's Random Trees Classication procedure, be sure to Normalize Input Data, to set Number of weak learners to 20, to set the Number of randomly selected features to 3, and to set the Minimum # records in a terminal node to 100.

a. What is the most important variable in terms of reducing the classification error of the ensemble?

b. For the default cutoff value of 0.5, compare the overall error rate, Class 1 error rate, and Class 0 error rate of the random trees on the test set to the corresponding measures of the single best-pruned tree from Problem 5.