What range of values of the relative sum of sqaures error is acceptable for a good neural network? I am getting around 0.9 of the relative error for 1 model and around 0.4 for another. Are both the networks significant?

  • $\begingroup$ Is the network being used for classification or regression ? $\endgroup$ – image_doctor Jun 28 '15 at 11:04
  • $\begingroup$ It is being used for regression! $\endgroup$ – Yash Lundia Jun 29 '15 at 12:23
  • $\begingroup$ Are you happy with both those levels of error for estimating your target function ? $\endgroup$ – image_doctor Jun 29 '15 at 12:32
  • $\begingroup$ Thats the problem! I do not know the significance of the term 'relative sum of squares error'. Are these errors out of 1? Meaning, does 0.9 mean 90% error? Or is it some other way? Please tell me the significance of the term 'relative sum of squares error'. $\endgroup$ – Yash Lundia Jun 29 '15 at 13:12
  • $\begingroup$ Also, is there some way to calculate RMSE in SPSS Neural Network analysis? Like instead of relative sum of squares error i want to calculate the RMSE of the network. $\endgroup$ – Yash Lundia Jun 29 '15 at 13:15

Check this:

A value closer to 0 indicates that the model has a smaller random error component, and that the fit will be more useful for prediction.

| improve this answer | |

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.