[ENH] Adding control over the `score` method of estimators about aeon HOT 2 CLOSED

I think its best if we add parameter string to the base class score function, defaulted to R2 for example in regression, accuracy in classification etc. and depending on that argument, the score function calculates a metric.

So, what are the scoring metrics we are planning to support for classification and regression respectively? @hadifawaz1999

from aeon.

I think its best if we add parameter string to the base class score function, defaulted to R2 for example in regression, accuracy in classification etc. and depending on that argument, the score function calculates a metric.

So, what are the scoring metrics we are planning to support for classification and regression respectively? @hadifawaz1999

For classification it should include the performance measures: accuracy, AUROC, Negative log likelihood (NLL) and balanced accuracy. These are the ones used in Bake off redux.

from aeon.