Open mrgloom opened 5 years ago
@e-sarkis To my knowledge, most of the methods used mean error for evaluation rather than distribution comparison. Maybe because it is more direct than distribution. Wondering how to guarantee that two same error distributions also have a very small mean error per sample?
In paper we have
Mean error
for model comparison, but I wonder is it common to use some statistical test to compare models? i.e. check that error distribution of model A and B is normal and then use some statistical test to compare distributions?