The first set of questions should be answered using the churn_train.arff ONLY. Open the data in WEKA and find the answers to the below questions!

Include answers to the following:

a) What is the baserate (percentage of people who churn) of the task (there are a number of different ways to provide the correct answer just be specific)? How would you characterize it (balanced vs. very skewed)?

b) What is the dimensionality of the features? In relative terms - do you think it is high or low?

c) What is the majority class (which is more common, churn or no churn)?

d) What accuracy would you expect to get if you always predict the majority class?