Reading for Gradient Boosting Machine


Basically, we'll read a series of three papers in order to get a complete understanding of gbm.  The sequence follows the temporal order in which these papers were published and will allow us to follow the development of the idea.  In order, here are the papers. 

1.  additiveLogisticRegression-Boosting.pdf

2.  greedyFunctionApprox-2001.pdf

3.  stochasticGradientBoosting-2002.pdf


Also look at the gbm package guide gbmPackageGuide.pdf  This has got some general explanation of gbm in addition to descriptions of the packages features and options. 


Here's some introductory code Ensemble.R

Here's R-code for the in class iris example:  gbm.R

Here's gbm set up to classify the sonar data  clemGBM.R


Recorded Lecture:

Part 1.


Part 2.