An old (from 2018) project on prediction of price of bitcoin and etherium.
Prediction of price of a cryptocurrency is a task complicated both conceptually and practically. A comprehensive multi-features dataset was used to preduct prices of bitcoin and etherium. A core element of the project was a comprison between "old" statistical methods (ARIMA) and Deep Learning (LSTM). Although their results were comparable, the former required a great deal of data manipulation which impeded thorough interpration of results. Both methods, however, yielded good short-term predictions.