Authors: Arnaud Stiegler and Redouane Dziri
The overall goal of this project is to predict whether a payment by a company to a medical doctor or facility was made as part of a research project or not.
We build the dataset from OpenPayments data, extract features, apply some preprocessing steps, fit a baseline model, do more feature engineering, build a more complex model, do some feature selection and finally build an interpretable model using all the insights gained from the previous steps.
The data used can be downloaded here: https://www.cms.gov/OpenPayments/Explore-the-Data/Dataset-Downloads.html (we used the 2017 data)