Stat Mills

Sometimes in my Data Science projects I need a quick and easy way to measure and visualize a predictive trend for several independent variables. My go to way to do this is to use Weight of Evidence and Information Value. Recently I put together some code in python to calculate... [Read More]

In my last post I covered how you can fit Penalized Splines using the glum library in Python. Notionally glum was built to fit Generalized Linear Models. However it was designed to give the user the option to pass in a custom penalty matrix. We took advantage of this capability... [Read More]

I recently discovered a new python library for fitting GLMs called glum. This package is developed not only to be fast but also allows custom penalties for different terms in the model, as well as including your own penalty matrix in your model fitting. This could allow us to include... [Read More]

Recently I encountered an interesting paper from the facebook research team that outlines a method for using decision trees (specifically boosted trees) to create transformed data to be used as input to a final logistic regression. I thought this was really cool and wanted to try and recreate the method... [Read More]

In order to build a prediction model for next season I need to know how good each team was in the previous season(s). A decent and simple way to measure team quality is to use the Massey rankings, which basically just finds the points above average that each team contributes... [Read More]