Tidepool Diabetes Hackathon
Projects | | Links: Github Code Repository
In this Hackathon, each team was asked to build a single classification model, of a randomly assigned type, to try to predict the classes,type 1, type 2 diabetes. We had to work within the constraints and best practices of our model type (Logistic Regression) for the duration of this hackathon. The goal was to build, test and quantify the utility of your given model. Once a basic model is built, we looked at how we could best tune the parameters, tweak the data, engineer variables to improve the predictive utility.
