Inspiration
Cryptocurrency is taking the world by storm! This means we don’t have a solid VIX or other market variables to truly be able to predict price changes. This further implies extremely high volatility that makes investing in crypto a gamble, more so than other investments. Our goal is for our hack to predict prices based on historical market data and recompile new data to generate an accurate prediction (within normal standard deviations)
What it does
Creating a price predictor that displayed not only numerical data but graphical representations as well. Considerations: Finhub Push time:15 minutes Ability to access minute resolution at 500 datapoints*minute
How we built it
The following components were put together to help create the graphs and data displayed. The databases served as our data sources and react was used to bring it all together and display everything. Finhub Docker React Multiple databases for data Neural Network Pytorch JSON
Challenges we ran into
Updates not fast enough
Accomplishments that we're proud of
The graph looks really clean
What we learned
How to use flask and how to use neural networks
What's next for Real Time Price Prediction
In the future we plan to continue working on the accuracy of the predictions as well as being able to provide with a live second to second feature. With more data and advancement in cryptocurrency these trends will be more predictable and volatility should become more stable. Another possibility is to implement our hack on order markets seeing the nature of the hack is predicting future prices the applications to commodities is far and wide.
Log in or sign up for Devpost to join the conversation.