We are new to the hackathon experience. This project not only served as a way of exploring a new technology we've never seen before, but also as a means of solving real world problems for the public.
What it does
Diagnoses patients within a training data set as having one of four conditions, name: Diabetes and hypertension, an infection, obesity, or Dyslipidemia.
How we built it
We executed a series of get requests to get identification of all 574 patients in the training dataset, and used those id's to make another get-request for observation data of all patients. We then used python to 'scrape' through this json file and extract key information we needed to diagnose if the patient had any of the four conditions.
Challenges we ran into
Getting the API to 'work' (we had trouble getting the patient data in the first place because of numerous faulty Get requests we made), and parsing through patient json files with python to extract useful information about the patient (i.e. bmi, weight, height, etc.) for use in diagnoses.
Accomplishments that we're proud of
What we learned
How to use RESTful API, and manage large datasets.