Closed datawookie closed 6 years ago
Mmm, yes, the number is actually unrelated - I calculated it as the proportion of loan applicants who did not repay their loans in the training set (ie. 8%). I realise I was confusing some concepts here, and have reworked this step, plus a bit more in repsonse to your other comment too so that I can better evaluate the logistic regression model.
Hi!
How did you arrive at the response threshold of 0.0807? Is this some magic number? :)
Best regards, Andrew.