Question: A decision tree to understand whether a customer churns has an initial Gini coefficient of .48. When the analyst computes Gini coefficients for a possible
A decision tree to understand whether a customer churns has an initial Gini coefficient of .48. When the analyst computes Gini coefficients for a possible initial split the Gini coefficient for a split based on whether there has been a customer service contact is .26, while the Gini coefficient based on a gender split is .34 and the coefficient based on a customer tenure split (that is, a recent versus long-time customer) is .23. Which split should the analyst use? Explain why you make this recommendation and what the meaning of the Gini coefficient is in non-statistical language.
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
