This public repo represents work done for the Kaggle Bike Sharing Demand competition
Thus far, the dependencies are the Python SciPy stack, plus GraphLab-Create.
Additional forthcoming documentation as I branch out into more models.
Linear algebra, multivariable calculus, and knowedlge of basic linear least squares regression.
That's all that seems necessary for implementing Gradient Boosted Trees.
The sources at the end of this wiki article are a wealth of information.
http://en.wikipedia.org/wiki/Gradient_boosting
http://homes.cs.washington.edu/~tqchen/pdf/BoostedTree.pdf
Written by one of the authors of the algorithm our submission used.
This wiki page is to explore and learn about Gradient Boosted Trees Regression, which is the backbone of our first sumissions to this Kaggle competition. Let's add material here to learn and understand it, and let's showcase our exploration of other methods similarly.