ABSTRACT
Over the last few years, cryptocurrencies have turned into a worldwide wonder known to many people. Bitcoin and Ethereum are widely used at present in global financial markets and so far, have received a significant value improvement and market capitalization with Ethereum being more impervious to downgrading than the other cryptocurrencies. The main aim of this study is to ascertain the prediction accuracy of both Bitcoin and Ethereum currencies using historical data (blockchain data with cryptocurrency data). This is the first study of this kind that we are aware of that predicts cryptocurrency prices using a Deep learning algorithm and blockchain information (Bitcoin and Ethereum). To accomplish this task, the blockchain data (2015–2018) of both Bitcoin and Ethereum currencies were collected to enhance the security and the prediction rate. Further, the market price of Bitcoin and Ethereum currencies were retrieved online. The effectiveness of the prediction accuracy was investigated using Deep learning approach with crypto currency data and blockchain data. The results of data analysis showed that when the blockchain data were used together with Bitcoin and Ethereum prices, the prediction performance is high for both currencies. In addition, the comparison between Bitcoin and Ethereum revealed that Ethereum currency has the highest percentage of the prediction accuracy and the lowest error rate. Moreover, descriptive analysis was undertaken for blockchain attributes such as difficulty, hash rate, number of transactions, average block size and miner’s revenue. The blockchain data directly influenced the prediction accuracy of both currencies (i.e., Bitcoin and Ethereum). The Deep learning prediction approach was found to be very effective for analysing blockchain and cryptocurrency data set. The price prediction of the cryptocurrency market price is vital as Bitcoin and Ethereum are effective in the present financial market.
