[GitHub] [incubator-mxnet] marfago commented on issue #16568: Different (not uniform) behavior in RMSE, MSE, MAE

[GitBox <gi...@apache.org>]

marfago commented on issue #16568: Different (not uniform) behavior in RMSE,MSE,MAE
URL: https://github.com/apache/incubator-mxnet/issues/16568#issuecomment-544661258
 
 
   I think this is more a bug than a new feature, since different metrics are calculated with different methodology (mean of sample vs mean of means of samples) which does not make metrics conceptually interchangeable (if I want the calculate the global RMSE with the current implementation I should estimate all the predictions and then update the metric, while in the case of accuracy I can update it at every batch). My feeling is that Accuracy should reflect the way RMSE is calculated (mean of means). This should also be clarified a little bit in the documentation.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services