Sepsis Analysis - BLHP

Inspiration

We were interested to see the effects of data and statistics in the health industry.

What it does

Models the likelihood of having a positive sepsis status.

How we built it

Used SQL to query data and python to clean, preprocess, and visualize our data.

Challenges we ran into

Runtime issues, had to reduce number of samples to be reasonable.

Accomplishments that we're proud of

Made 7 models that had varying accuracies. Achieved a 79% accuracy score.

What we learned

Feature importance, furthered our SQL and Python knowledge.

What's next for Sepsis Analysis - BLHP

Using the whole dataset, more features, more complicated models.

Built With

• python
• sql