How, if at all, can one compare the "fit" of a simple linear vs. non-linear regression model to observed data?

I apologize if I didn't search long/hard enough for the answer, but I cannot find anything concrete.


  • $\begingroup$ What software are you using? $\endgroup$
    – Michelle
    Jan 26, 2012 at 23:39
  • 2
    $\begingroup$ Can you share your purpose in comparing the fit of the different models? To develop a predictive model? To describe a particular data set? To provide evidence for or against a particular theory? You might choose different comparison approaches depending on your purpose. $\endgroup$
    – Anne Z.
    Jan 27, 2012 at 3:05
  • $\begingroup$ Anne, at this point it is simply for description. I would like to quantitatively be able to say that a non-linear fit is preferred. $\endgroup$
    – Patrick
    Jan 28, 2012 at 17:21
  • $\begingroup$ Michelle, I use R. $\endgroup$
    – Patrick
    Feb 3, 2012 at 17:20
  • $\begingroup$ Would using AIC/BIC or the LogLik be sound statistically to make a choice between non-linear and linear models on the same dataset? $\endgroup$
    – user9974
    Mar 19, 2012 at 23:33

1 Answer 1


Perform cross-validation on each model on a development set to find the best hyper-parameters / accuracy estimate for each. Then check that the accuracy estimates are still valid on some held out data that you did not use during the model selection / parameter tuning phase.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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