Bitcoin Price Analysis With Deep Learning Models

Abstract

Cryptography is used by cryptocurrencies to safeguard and authenticate transactions, as well as to regulate the creation of new units. Cryptocurrencies, as opposed to traditional money, which is often issued by a central authority like a government are decentralized meaning they are not managed by any one entity or group. One of the most well cryptocurrency is "Bitcoin". Cryptography is used by Bitcoin, a decentralized digital currency, to safeguard transactions and regulate the generation of new units. It is exchanged on different cryptocurrency exchanges and runs without the help of a central bank. However, the future currency is the ultimate digital currency. In this paper we are going to inculcate with the neural network models like sequence-to-sequence, ARIMA(Auto Regressive Integrated Moving Average), LSTM(Long Short-Term Model). This process is certainly made to analyze the bitcoin price over the past couple of years. We trained our model accordingly and found out 96.21% accuracy for the sequence-to-sequence and 94.2% for the ARIMA as well. Eventually this paper and its work insinuate regarding the highly accurate performing algorithm and its analysis as well.