In classifiers overview: fixes and updates for scikit-learn developments

• mention in boosting: Histogram-based Gradient Boosting Classification/Regression Tree: HistGradientBoostingClassifier/Regressor; this estimator is much faster than GradientBoostingClassifier/Regressor for big datasets (n_samples >= 10 000).
• update classifiers info on DT pruning (add example?): tree pruning based on minimal cost-complexity via ccp_alpha=0.05 kwarg
• rm n_jobs if used in KMeans estimator (re-worked - significantly faster and more stable, but also uses OpenMP so n_jobs has no effect (instead need to set OMP_NUM_THREADS=... env var)
• mention stacking in ensemble methods
  • StackingClassifier and StackingRegressor: a stack of estimators with a final classifier or a regressor. Stacking allows to use the strength of each individual estimator by using their output as input of a final estimator. Base estimators are fitted on the full X while the final estimator is trained using cross-validated predictions of the base estimators using cross_val_predict.
  • good overview/intro: https://towardsdatascience.com/ensemble-methods-bagging-boosting-and-stacking-c9214a10a205

changed milestone to %Workshop 3rd edition (2020-2021)

created merge request !29 (merged) to address this issue

mentioned in merge request !29 (merged)

marked the checklist item rm n_jobs if used in KMeans estimator (re-worked - significantly faster and more stable, but also uses OpenMP so n_jobs has no effect (instead need to set OMP_NUM_THREADS=... env var) as completed

marked the checklist item update classifiers info on DT pruning (add example?): tree pruning based on minimal cost-complexity via ccp_alpha=0.05 kwarg as completed

marked the checklist item mention in boosting: Histogram-based Gradient Boosting Classification/Regression Tree: as completed

marked the checklist item mention stacking in ensemble methods as completed

closed with merge request !29 (merged)

mentioned in commit e8efab3c