Dear organizers!
What is the motivation for choosing such metrics, as Paerson correlation, etc, when there are commonly used, classic metrics for regressor quality estimations - explained variance score, mean squared error, r2 score...?
I tried myself to compare the modelling results by classic metrics and by metrics provided - and the ranking of different regression models with two approaches is different.
Dear organizers!
What is the motivation for choosing such metrics, as Paerson correlation, etc, when there are commonly used, classic metrics for regressor quality estimations - explained variance score, mean squared error, r2 score...?
I tried myself to compare the modelling results by classic metrics and by metrics provided - and the ranking of different regression models with two approaches is different.