Random forests, SVMs and Linear Models in the cryptocurrency space
This analysis looks at the predictability of 3 major cryptocurrencies: Bitcoin, Ethereum and Litecoin through standard ML algorithms.
Introduction
The full notebook is available here: https://github.com/IanLDias/algo-trading/blob/main/notebooks/crypto/ML%20approaches%20to%20crypto.ipynb
This is part of a broader project available here: https://github.com/IanLDias/algo-trading
Data Sources:
Two main data sources were used in this analysis, one to collect data on historical prices and the other to collect relevant IVs.
- Coinmetrics: 10 IV's regarding blockchain data
- Transaction information (count, size, fees, difficulty)
- Market cap
- Active addresses
- Cryptocompare
- Historical prices (OHLCV)