Inspiration

Recently, the variety of assets that can be invested has skyrocketed, mainly in terms of Crypto. However, due to the extreme volatility of cryptocurrency, many investors end up liquidating and losing all their money. Therefore, the possible prediction on cryptocurrency has a huge impact both socially and commercially.

What it does and How we built it

By using ARIMA and LSTM models presented in this report, both retail and institutional investors can have a more reasonable buying and selling time and maximize their profits. Without having any additional expertise about Bitcoins, investors can also have equivalent accurate predictions through our models.

Challenges we ran into

The time taken to train the model is much longer than expected and there is a lack of knowledge of time series analysis.

What we learned

Through the exploration of the modelling, we learn about basic models of time series analysis and some criteria for choosing the proper hyper-parameters.

What's next for Prediction of Bitcoin Price

  1. Comparison with other cryptocurrencies to find the potential relationship.
  2. Using all time data to predict

Built With

Share this project:

Updates