glouppe/tutorial-pydata2014

IPython notebook for PyData SF 2014 tutorial: "Gradient Boosted Regression Trees in scikit-learn"

PyData 2014: Gradient Boosted Regression Trees in scikit-learn

Instructor: Peter Prettenhofer

• email: peter.prettenhofer@gmail.com
• twitter: @pprett
• github: pprett

This repository contains the tutorial material for the Gradient Boosted Regression Trees in scikit-learn tutorial at PyData 2014.

Requirements

This tutorial requires scikit-learn 0.14+. I strongly recommend that you update to the current master version (to be 0.15) since it includes a significant speedup for our tree ensembles including Gradient Boosted Regression Trees. You might get slightly different results (Cal-housing scores & variable importances) depending on your version of scikit-learn.

Notebook Static View

• http://nbviewer.ipython.org/github/pprett/pydata-gbrt-tutorial/blob/master/gbrt-tutorial.ipynb