From the course: Advanced Predictive Modeling: Mastering Ensembles and Metamodeling

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Gradient boosting demo

Gradient boosting demo - SPSS Tutorial

From the course: Advanced Predictive Modeling: Mastering Ensembles and Metamodeling

Start my 1-month free trial

Gradient boosting demo

“

- [Instructor] What I'm about the show you is not the typical way that folks explain gradient boosting. Let me explain. Gradient boosting is a different attempt to give wrong answers added weight. In add a boost, you weight the rows. The rows that you get wrong get more weight. Here we're going to use a different approach. What we're going to do is calculate the residual, which as you probably know is observed minus expected, which in this case is the actual miles per gallon minus the expected value that I got from building a very simple regression model. What I'm going to do now is build another model that makes the residual the dependent, that makes it the target variable. Now, when you read or learn about gradient boosting, they're going to talk about loss functions and they're trying to minimize the loss function. The residual here is our loss function. That's our error and we want to make it as small as possible. So I'm going to go ahead and build a tree now where my goal is not…

Contents