Fundamentals of Predictive Analytics with JMP, Second Edition

Chapter 13: Bootstrap Forests and Boosted Trees

Introduction

Bootstrap Forests

Understand Bagged Trees

Perform a Bootstrap Forest

Perform a Bootstrap Forest for Regression Trees

Boosted Trees

Understand Boosting

Perform Boosting

Perform a Boosted Tree for Regression Trees

Use Validation and Training Samples

Exercises

Introduction

Decision Trees, discussed in Chapter 10, are easy to comprehend, easy to explain, can handle qualitative variables without the need for dummy variables, and (as long as the tree isn’t too large) are easily interpreted. Despite all these advantages, trees suffer from one grievous problem: They are unstable.

In this context, unstable means that a small change in the input can cause a large change in the output. For example, ...