I want to build a model to support decision making in order to propose or not loan insurance to clients. Because sometimes clients asking loan and loan insurance have less chance to have their loan accepted by a bank and sometimes more chances.

There are three actors in the problem: a bank, a loaner applicant (someone who ask for a loan) and a counselor. The counselor studies the loaner application and if it has a good profile it will propose to him loan from banks that fits his profile. Then the application is sent to the bank but the bank could refuse the applicant (based on criteria we don't know).

The counselor has also to decide whether or not he will propose to the loaner applicant a loan insurance. I want to build a model for that decision.

The risk is that some banks reject loan applicant who accepts a loan insurance and other banks accept more applicants with a loan insurance. But there aren't rules regarding banks since some banks accept or reject applicants with loan insurance according of the profile of the applicants and the type of acquisition.

Thus, the profile of the applicant and the bank he is applying to can matter in their rejection from banks but all criteria influencing the decision are quite uncertain.

Even though it is a classification problem, in my dataset I don't have a good label for loan insurance proposal. I have a features that says if the insurance was proposed or not and to less than 1% the insurance was proposed. I have a label that says if the clients application for a loan was accepted or not.

Thus, the data I have is former applicants profile - and banks who propose loan according to what the applicant wants - and if they were accepted or not by the bank and if they wanted a loan insurance or not.

I thought of combining the label with the information but I don't really know how or maybe doing multi-label classification but also I don't really know if it does fit to the problem.


Maybe I didnt understand it correctly but it doesnt seem like a multi-class problem. I think you can just model to know about the final status of the loan application and test/use the insurance information you have as a dependent variable. Just a guess, but given the low fill rate, it might not come up as an important variable.

  • $\begingroup$ Do you mean predicting the final status with and without the insurance and compare the probability of the outcome ? $\endgroup$ – Minila S Oct 3 '18 at 12:05
  • $\begingroup$ No I mean just model the final status and use the insurance information as a 1-0 flag field in the model. $\endgroup$ – Dan9ie Oct 10 '18 at 10:54

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.