1

I am trying to learn decision trees but it has been difficult because the examples are extremely long and tedious and everybody seems to have a different algorithm in mind

enter image description here

After some digging I found a reliable set of notes online. However, I have the following two questions

  1. What algorithm is this?

  2. What is the meaning of $\mathbb{I}$?

It would be great if anyone can clarify this.

Here are the notes: http://www.stats.ox.ac.uk/~flaxman/HT17_lecture13.pdf

Norman
  • 297
  • 2
  • 11

1 Answers1

1

The tree process exposed here is the CART algorithm. Another link covering it.

The II is an indicator function. It basically means 1 if xxx else 0.

Beta is the probability of going either way compared to the splitting parameter

Mayeul sgc
  • 205
  • 1
  • 7