QuantFrog

Inspiration

We were inspired to create this project because we saw a need to make quantitative trading analysis more accessible. And throughout the development of this project, we experienced how difficult it is to really get software deployed.

What it does

Utilizes historical financial statistics for Fortune 500 companies to predict future Earnings Per Share

How we built it

We ended up building the project using React for the easier Frontend development. For the backend, we used Python and Jupiter Notebook to webscrape for data, build the data frame, and use it to train our predictive model. We also used webscraping to get data for our Sentiment Analysis, which was done on Google Colab

Challenges we ran into

Webscraping was one of the biggest challenges we encountered. None of us had a lot of experience and given that a lot of the available information was protected from public use, it was very difficult to find a way around our data shortage.

Accomplishments that we're proud of

Finding a way to collect proper data despite our unfamiliarity. We were really happy with the data we were able to get and we're proud of the performance of the model.

What we learned

We got better with creating models and we learned different ways to get useful data. This was also a new experience with Sentiment Analysis and integrating it with float parameters.