Greenway Health hackathon ยท 2nd place
Risk Stratification
An internal hackathon project at Greenway Health. My team took the data points that exist in an electronic health record and trained a Random Forest classifier to predict a patient risk score and classify patients. It placed 2nd, and the CTO presented the award.
The patient
Care starts with knowing who needs it first
Clinics see many patients. Knowing which ones carry the highest risk helps care teams decide where to spend their attention.
The records
Data points from an EHR
The team used the kinds of data points an EHR holds and stored them with SQLite, working in Python.
The forest
A Random Forest classifier
Many decision trees vote on each patient. The model predicts the risk score from the given data points.
The groups
Classifying patients
The output sorts patients into groups by risk, from lower to higher, so the highest-risk patients stand out.
The score
A risk score for every patient
The idea scored 2nd place in the Greenway Health internal hackathon and was awarded by the CTO. More detail will be added here.
Placeholder: demo video or screenshots