What statistical or machine learning methods do companies like Lending Club use to segment their customer base into loan grades A1-G5? What would a reasonable partitioning method look like after running, say, a simple regression?

  • 1
    $\begingroup$ Did you try to Google 'credit risk modelling' because you can use a whole bunch of techniques, a.o. Logistic regression, neural networks, support vector machines, survival analysis, random forests, boosting, ... $\endgroup$ – user83346 Feb 8 '17 at 5:49
  • $\begingroup$ @fcop There are tons of guides on getting relevant output but not how to properly segment that output. For example, how can probability of default be segmented into loan grades? $\endgroup$ – Test Feb 9 '17 at 14:34
  • 1
    $\begingroup$ What about ordered logistic regression e.g ? That gives you a score and the thresholds to use to partition the score range into rating classes ? $\endgroup$ – user83346 Feb 9 '17 at 15:45

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.