Into the Ether

Inspiration

Cryptocurrency is revolutionizing the world of finance. Our group shared an interest in the intersection of finance and technology so we set out to predict the price of the Ether coin using machine learning.

What it does

The project using historical market data from the Gemini Exchange, a cryptocurrency exchange that allows user to buy and sell various coins such as Bitcoin and Ethereum. Using a Recurrent Neural Network, we predicted the price of Ethereum in the future.

How we built it

We used various Python packages such as sklearn, numpy, pandas, and keras to implement machine learning algorithms such as a linear regressor and a long short-term memory recurrent neural network.

Challenges we ran into

None of the team had used a recurrent neural network before so we had to search up a tutorial on how to do it. We followed the steps, understood the procedure, and were able to make it work very well.

Accomplishments that we're proud of

We were very suprised at the accuracy of the neural network and were astonished that it was able to predict a general trend not many people saw coming (the past year's cryptocurrency boom).

What we learned

We learned that it is important to scale our features for the recurrent neural network and that a recurrent neural network is a very useful tool for predicting values with chronological data.

What's next for Into the Ether

We hope to continue advancing the model to begin forecasting the future prices and implementing the Gemini Exchange's API so our model can begin trading with a small amount of money such as a $100.