Consider the same dataset as in the previous question. Now the retailer $($data analyst$)$ wants to use$$ Decision Trees to classify new customers.

$2\mathrm{.}2\mathrm{.}2$ Practical assignments $35$

$($a$)$ What is the classification error rate for attribute A$?$

$($b$)$ What is the classification error rate for attribute B$?$

$($c$)$ What will be the splitting attribute in the top $($root$)$ of the Decision Tree if one uses the classification

error rate?

$($d$)$ What is the Gini index for attribute A$?$

$($e$)$ What is the Gini index for attribute B$?$

$($f$)$ What will be the splitting attribute in the top $($root$)$ of the Decision Tree if one uses the Gini index?

$($g$)$ Construct the full Decision Tree, using the error rate as heuristic, and what is the overall classification

error rate on the above dataset?

$($h$)$ Is this classification error rate an optimistic or pessimistic estimate of the error rate on unseen new

data? Explain your answer