Comparing differences in RMSE

Asked Oct 25 '19 at 04:01

Active Oct 25 '19 at 04:01

Viewed 40 times

I tried to search on the web as well as here. Seems like no answer. There was a question previously asked in stack exchange, but it was discussed instead of being answered.

I have RMSE from 6 groups. How do i test for significant differences in RMSE? Is T-test, ANOVA or Wilcoxon test okay? Some argue that is does not make sense.

If RMSE cannot be compared using such test, how can i make a decision as to whether there are any significant differences?

Thank you.

asked Oct 25 '19 at 04:01

Candice

Could you link the question you mention? – Richard Hardy Oct 25 '19 at 05:55
@Candice: You may like to see [this](https://stats.stackexchange.com/questions/253613/diebold-mariano-with-rmse-as-loss-function). – Dayne Oct 25 '19 at 07:32

Comparing differences in RMSE

0 Answers0