Tree-Based Machine Learning

Here ML means:

• Magic Leap

• Machine Learning

• Much Love

Where is the root of a tree located?

• Up the tree

• Down the tree

Regression is:

• Hypnotic therapy

• Linear modeling of numeric label

• Features-numeric label modeling

Impurity measures how heterogenous is the label:

• within a set

• comparing different sets

What SHAP contributes compared to measures of feature importance is:

• nothing

• size (and sign) estimates

• global and local contribution estimates

The most robust method is:

• RandomForest

• XGBoost

• Decision tree

Which one needs less trees?

• RandomForest

• XGBoost

• LightGBM

Which one has more randomness?

• XGBoost

• RandomForest

• ExtraTrees

What docker contributes compared to PICKLE+FLASK is:

• going outside developer laptop

• docker image of the model

• visibility and usability