Comparative study of vector autoregression and recurrent neural network applied to bitcoin forecasting
Zakariae El-Abdelouarti Alouaret · 2017
In this thesis we are exploring the prediction of next day Bitcoin (BTC) price through the usage of Recurrent Neural Networks (RNN) . Our aim is, by using state-of-the-art techniques, to predict the price of BTC with higher accuracy than the previous works in the literature. This thesis uses up to 27 time series, spanning from 03/01/2009 to 28/04/2016, with a granularity of 1 data point per day.