Introduction to Tree-Based Machine Learning - Section 1: Regression

Table of Contents: click the button to the left of the full screen button (hover your mouse over the lower right hand corner of the video)

Learn the basics of Salford Systems' data mining software in one-hour video overviews. This is a great place to start if you're new to to decision trees. Video presentations include introductions to CART, MARS, TreeNet, Random Forests, and the basics of data mining.

Introduction to CART Decision Trees for Regression

This video provides an introduction to the methodology underlying CART® software in the context of regression (i.e. quantitative target).

How to Write a Decision Tree (CART®) as an Equation

Learn how to express any CART tree mathematically. Writing the equation of a CART tree will help you understand how linear effects, nonlinear effects, and interaction terms are handled in CART. Understanding the equation will also provide insight into advanced machine learning techniques where CART is the foundation such as TreeNet Gradient Boosting, RandomForests, MARS® regression splines, ISLE™ model compression, and RuleLearner™.

How are Variable Interactions Modeled in Decision Trees (CART®)?

CART automatically models variable interactions. Since CART is the foundation of Random Forests and Gradient Boosted Trees, both of these methods also automatically model variable interactions. In this video we discuss how CART models variable interactions locally by expressing CART as an equation. Prior to watching this video, it is strongly recommended that you first watch the video “How to Write a Decision Tree (CART) as an Equation”

Introduction to Random Forests for Regression

This video provides an introduction to the methodology underlying RandomForests® software in the context of regression (quantitative target). Prior to viewing this video please first watch the video Introduction to CART Decision Trees for Regression because CART decision trees form the foundation of the Random Forest algorithm.

Introduction to Stochastic Gradient Boosting for Regression

This video provides an introduction to the underlying methodology for TreeNet® software in the context of regression (i.e. quantitative target). Prior to watching this video it is recommended that you first watch the video Introduction to CART Decision Trees for Regression because CART decision trees form the foundation for stochastic gradient boosting.

Introduction to MARS Nonlinear Regression Splines

This video provides an introduction to Multivariate Adaptive Regression Splines. Prior to watching this video, it is recommended that you first watch the video "Introduction to CART decision Trees for Regression"
Table of Contents: click the button to the left of the full screen button (hover your mouse over the lower right hand corner of the video)