# Missing Values in Classification

I'm working on a classification problem. I'm trying to build a model which can predict if a bank client will get a loan or not. Some of clients have co-borrower and the majority don't.

I also have information on co-borrowers like salary, etc. but as the majority of clients don't have co-borrower, I have missing values. How can I impute this missing data ?

thank you

• Depends on your whole data. But one simple way, just drop co-borrowwer – bakka Oct 8 '18 at 9:43
• thank you sir for your response unfortunately, i can't drop co-borrowers because in the case when they exist , client get the loan so I think it's a pertinent information. – Haythem Mzoughi Oct 8 '18 at 9:46
• How about just a binary has_coborrower indicator? – shadowtalker Oct 8 '18 at 15:25