We know how stressful dealing with financial decisions, especially when taking out loans is for most people. Although we have large datasets giving us information on how an individual repays their long/short term loan, extracting useful data and conclusions can be a hard task. This was why we challenged ourselves to try and make sense of one example dataset in order to guide individuals to make more responsible, well-rounded decisions.
What it does
The webpage extracts multiple fields from four datasets made available by Fannie Mae. It visualizes the data based on some categories such as FICO score, State, Term of the loan and etc.
How We built it
We first wrote a python script in order to clean the dataset given to us by Fannie Mae. Secondly, we tried to filter some categories in order to find patterns between the data fields. We created a database using the categories that we were more interested in, using SQL. After finding some important risk factors that usually lead borrowers to underperform in their payments, we started creating an interactive webpage using HTML and CSS that first gives the user two main charts -extracted from the given dataset- then asks the user some questions and filters out the cleaned dataset in order to provide customized visualization for that specific user. We then deployed our webpage using docker on Google cloud servers.
Challenges We ran into
We learned a lot on the way since we did not know enough about anything we used for this project.
Accomplishments that We're proud of
Staying awake for the entire night and completing what we had planned to build.
What We learned
With enough time energy and motivation, we can learn incredibly fast.
What's next for You're not A-Loan
Extracting data from other large datasets and try to complete our database. Currently we do not have enough information to always provide precise statistics. For example, we do not have data points for every single FICO score, or we do not have enough data points for some states.