This project formulate neural network models that utilize the LSTM (Long Short Term Memory) algorithm, one of the variants of Recurrent Neural Network Algorithm to predict the closing price value and the probability of rise or fall of BTC for the next hourly and daily timestamps with regression and classification methods