Inspiration 🌟

We wanted to try out a new programming language not typically seen in hackathons: R. Usually used in statistics for academia and sometimes industry applications, R provides a data science dashboard framework called shiny.

What it does 🌟

Most stock trading platforms have a TON of metrics and quantitative data that inexperienced users may find overwhelming. Furthermore, These platforms are not too personalizable. Therefore, FinCompute seeks to combine a few easy-to-understand charts and tables for any selected stock and time period while also giving users some personal finance attention with a customizable financial goals list for 2022. Finally, a pre-trained deep learning model called finBERT (based on the famous BERT model) can process sentiment analysis on financial news, articles, and other textual materials. This adds qualitative details in addition to quantitative data helpful to make stock trading decisions.

How we built it 🌟

We used both R and Python, although R was predominant given the usage of shiny. We used yfinance in Python to fetch financial data directly from Yahoo Finance and employed plotly in R

Challenges we ran into 🌟

Using R on such a major task.

Accomplishments that we're proud of 🌟

Getting the app running locally.

What we learned 🌟

  • Using shiny to build UI and reactive components
  • Interpreting Python in an R environment with reticulate
  • Metrics on stocks

What's next for FinCompute 🌟

We hope to leverage either Tensorflow or Keras (in R) to train a custom LSTM model for stock price prediction based on a recent window of price changes.

Built With

Share this project:

Updates