Inspiration: At the starting of the competition, we really didn't have a project idea so we began by browsing through the hackathon sponsors and their award criteria. We then stumbled upon the Fannie Mae Challenge to predict the number of home sales in certain communities. However, we didn't want to make an application that solely benefits corporations but something that socially impacts the community as well. We realized that on top of the buying predictions we could incorporate a representative map of Loan Acquisition concentrations in various states across the nation. Furthermore, we wanted to create a second portion of our web application that allows people to establish "safe houses" in numerous communities so people would be able to safely navigate through possibly unsafe areas.
What it does: Our application successfully predicts national housing sales. Furthermore, it allows users to create their own accounts to view their address information and Loan Acquisitions in their areas. Also, by using Zillow API given an address and
How we built it: The application utilizes HTML and CSS for the front-end portion of our web application to display our findings and create a friendly User-Interface. For the account authorizations, we are utilizing the Firebase API, which allows users to register accounts with their emails which is then transported to the Firebase User Authentication database. For the back-end data analysis and Machine Learning Algorithm, we are using python. Our application utilizes the MapBox API for displaying our interactive map. Then, the Fannie Mae API was used to acquire national housing which was then analyzed through the Machine Learning Support Vector Machine Algorithm.
Challenges we ran into: The "Safe House" aspect that we wanted to implement was only partially configured because we were able to successfully acquire user information through Firebase API and then we were able to develop the map as an extension to our webpage.
Accomplishments that we're proud of: We were able to successfully predict the housing sales through our Support Vector Algorithm. This data was also successfully represented through our Interactive MapBox United States Map that depicted Loan Acquisitions which is directly correlated to house sales.
What we learned: We became well-versed with utilizing the Firebase Authentication software allowing us to create User Accounts in a reliable and encrypted manner. Also, our utilization of the Support Vector Machine developed our understanding of the capabilities of Machine Learning and all of the possibilities. We gained experience using the MapBox API, which we had no experience with before.
What's next for Hack TJ 2019: Attempt to implement a chatting system in our web application along with the safe haven aspect. Furthermore, attempt to improve our User-Interface to make it more streamlined.
Log in or sign up for Devpost to join the conversation.