In the data science world, I have always evaluated the performance of logistic regression models simply using ROC/AUC. However recently, I've read from some traditional statistics source about some measures of a logistic regression, including:

for goodness-of-fit: Pearson statistic, Deviance, and Hosmer-Lemeshow statistic; for quality of prediction: Nagelkerke's $R^2$, Concordance measure.

So, which metrics are better for what scenario? Also, I'm pretty confused how goodness-of-fit is different from predictability?


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.