NoFinger Inc.

What it does

We used anonymous patient information in order to predict if they were going to require their drug for free.

How we built it

We selected our features based off correlation to patient use of free drug and visualized proportion of free drug use within each feature. These features were used to fit a logistic regression model or build a neural network.

Challenges we ran into

Our main challenge was working with categorical variables and choosing which features to use in our models.

Accomplishments that we're proud of/What we learned

We built a model and learned a lot!