Can handle…

• … all types of data (discrete and continuous)
• … all types of relationships among the independent variables
• … all kinds of relationships with the dependent variable

Fastai tutorial shows how to use classifier with sklearn.

Example

Dataset

Linear Regression doesn’t work here:

Find the root

Which of the 3 independent variable should be on the top of the tree? What is the best split?

Leaves that contain mixtures of classifications are called Impure.

Above we only see one Leaf which is Impure. It looks like Loves Soda does a better job in classifying who loves the movie. Is there a way to quantify the difference between Loves Soda and Loves Popcorn?