Hyperparameter tuning but without overfitting the scikit learn models, hence the naming "naive".
Corresponds to a case with sufficiently large datasets for not caring about overfitting.

budget150
budget500
datasetartificial
datasetartificialcos
datasetartificialsquare
datasetauto-mpg
datasetboston
datasetdiabetes
datasetred-wine
datasetwhite-wine
depth
dimension1
dimension11
dimension2
dimension3
dimension5
dimensionnan
regressormlp
tree
workers1
workers125
workers37
