Cryptonite uses ml and numerous apis
to generate its daily market prediction.

Through the analysis of people's tweets, we were able to develop a model that predicts how the price of Ethereum will change from day to day.

7 days of
Ethereum

The graph to the left displays the price of Ethereum over the last seven days (in USD). Due to its clear volatility, we believed its market swings to be heavily influenced by hope and fear.

7 days of
human emotion

To test our theory, we produced this graph - seven days of Ethereum tweet sentiments on a scale from 0-1 (with 0 being entirely negative, and 1 being entirely positive). The correlation with price is very strong, as R^2 is greater than 0.44. This suggests that tweet sentiment is a strong predictor of Ethereum pricing.

How does it work?

Our project relies on the advanced capabilities of a number of web services to function.
Interested in learning more? Read below to understand each step of the process.

  • Market Prediction

    Cryptonite anticipates market trends through the use of an advanced machine learning algorithm.

  • Dynamic Site

    Cryptonite employs a Django back end to dynamically update as predictions change.

  • 2-class SVM

    Cryptonite uses an SVM built with Azure's Machine Learning Studio to identify tweet sentiment.

  • Private API

    Cryptonite uses a private API powered by Azure's Web Service to access data efficiently.

  • Crypto Tracking

    Cryptonite accesses the World Coin Index's API to record the price of Ethereum over the last 90 days.

  • Accessible Anywhere

    Cryptonite is available in the entire world through the power of Heroku's free hosting plan.