I have a few questions about data leaks. Particularly, I'm interested in a credit scoring data can have leakages. I'm at the stage of data acquisition and I suppose I have target leak but not sure.

So, I'm trying to build a credit scoring model to predict if the customer will make a request for a loan. The target variable is binary and is encoded in the following way: 1 if the customer had a loan before and 0 if not. (0 can be an insurance product but not loan. Still, that customer is a bank's client.). I have one feature, named "Loan type" which contains the following values: a mortgage loan, consumer loan, and insurance. mortgage and consumer loan are 1 in the target variable and insurance is 0 in the target variable.

Is this feature source of target leak or any data contamination? if it is how to deal with it and why it is data leak source? If it is not data leak source why we can consider such a way?

I did not do any feature engineering or partitioning yet.

I struggle to find answers to my questions. Any help will be appreciated.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.