Project Pigpen

Inspiration

As a team, we wanted to enhance our skills in data science at an advanced level and delve into deep learning.

What it does

We currently have our PostgreSQL database set up. If we had more time, we would have implemented an RNN with Long Short-Term Memory to analyze underweight infants and infant mortality over time in Texas counties.

How we built it

After cleaning up the data files, we transferred them into the PostgreSQL database.

Challenges we ran into

Data parsing was a huge challenge, but we were able to overcome it. The largest challenge was developing our understanding of the challenge and the context of the data. We thought we understood it at first, but we eventually came to new realizations that were able to strengthen our views on what mattered and didn't mattered in the datasets.

Accomplishments that we're proud of

We are a very resilient team. We always went in with a plan. Whenever our plans would fall through due to new realizations, we were able to pick ourselves back up create a new, stronger plan.

What we learned

We learned just how important it was to go into great detail when analyzing the requirements of the challenge and the challenge's data. With having a time limit, we didn't take as much time as we should've at the beginning to truly understand things. However, we were able to make due with the time we had once we had a full grasp of the challenge.

What's next for Project Pigpen

Besides implementing the RNN, we would have liked to have a data visualization program to be able to select a county and see it's analysis of underweight infancy and infant mortality over time. We would also have liked a user to be able to compare a county's data to the data of other countries, the entire state of Texas, and other states.

Built With

• keras
• lstm
• pandas
• postgresql
• python
• rnn