Model Calibration, Ensemble and Stacking Models Addition

[ozguraslank]

Explanation:

Adding model calibration, ensemble and stacking models features to enchance outputs.

This feature should be either in wide experiment or included in a different function that user can call after the experiment finishes.

• Model Calibration: link
• Ensemble Learning scope: VotingClassifier for classification and VotingRegressor for regression
• Stacking scope: Not certain yet but It can be something like playing with top3 models to create combinations e.g. (1st+2nd) + (1st+3rd) + (2nd+3rd) models

Notes for Stacking:

Maybe a parameter decides Top X can be defined that results more model combinations and another one determines how many models will each combination include