UH_FDA

We knew nothing about Python and data science before this competition, and Datathon provide us with an opportunity to learn so much! Thank you to all the organizers and the mentors in Rice Datathon.

Comment

How we built it

We first cleaned and drop some columns that we did not use throughout our project as well as any null (NaN) variable. We also changed some column titles and changed some rows to lowercase for easier read and easier data processing. Then we divide our data into 2 datasets: "Suspect" and "Concomitant". We evaluated the severity of the product using the associated outcomes with severe outcomes (e.g. death, hospitalization, disability, etc) and assigning each report with one point stored in the Hospitalization column. We then add up the points in the hospitalization column, take the average, and have the severity score out of the scale of 3. We concluded that the substance that resulted in the worst consequence is Alcoholic Substances and the most common products in both the suspect products and concomitant products are the products in the Vitamin category.

Challenges we ran into

We don't have any experience in Python and data science prior to this event, we learned everything through the Beginner's Workshop hosted by Grace Wang, one of the organizers, as well as our mentor throughout the duration of the project.

Accomplishments that we're proud of

We learned so much more about Python, pandas, matplotlib, and just data science ideas in general.

What we learned

Python, pandas, matplotlib, and how to approach a set of data

What's next for UH_FDA

We look forward to participate in similar events to earn even more experience so that hopefully we can come back next year stronger!

Built With

Share this project:

Updates