Question 21
Consider a credit card company analyzing its customer base to assess the risk of default on credit card payments. The company wants to use the WoE technique to transform the "Income Level" feature, which is a categorical variable with values "Low," "Medium," and "High," based on the observed relationship with the target variable (default or non-default). After analyzing historical data, the company finds that the WoE values for each income level are as follows:
● WoE for "Low" income level = -0.75
● WoE for "Medium" income level = 0.15
● WoE for "High" income level = -0.90
What does the negative WoE value for the "High" income level suggest in the context of credit risk modeling?
Customers with a high income level are less likely to default on credit card payments.
Customers with a high income level are more likely to default on credit card payments.
The "High" income level has no significant impact on the risk of default.
WoE cannot be interpreted for categorical variables with more than two levels.