It is one of the most complex challenges faced by Champaign County and we noticed there was a lot of data surrounding it.
What it does
Given a list of locations and various features, such as average household income and English as a second language percentages, our algorithm determines the relative impact each feature has on poverty.
How we built it
Challenges we ran into
Front-end stack development in deploying our website turned out to be a major challenge. None of us had ever developed frontend, so this was a major learning experience.
Accomplishments that we are proud of
We used a unique model of data analysis that none of us had ever worked with.
What we learned
We gained experience creating and hosting a dynamic website from scratch.
What's next for CU-Poverty
We want to potentially host it on a cloud service such as Azure or AWS, and also train our model on a larger dataset to increase accuracy and extrapolation capabilities. Additionally, due to the small dataset, we could enhance and increase our dataset through various data augmentation techniques, such as the SMOTE algorithm.