Abstract:

Money and other payment instruments have undergone many changes until today. With the widespread use of the internet brought about by technological developments, today cryptocurrencies are preferred as an alternative to payment tools. The purpose of this study is to investigate the method that will make the future price prediction of cryptocurrencies as close to the truth. Thus, market analysis will be facilitated by contributing to the developments in payment instruments. During the study, artificial neural networks (ANN) and long short-term memory (LSTM) artificial intelligence methods were used. The results of both methods used in the study were evaluated separately and a performance comparison was made in line with the results produced. The three preferred cryptocurrencies in the study were determined as XRP, BNB and ETH. The data to be analyzed in this context was taken from the yahoo website and the data set covers the dates between 01.01.2020-07.05.2021. As a result of the comparison analysis, it has been revealed that the estimation results are successful and meaningful in the matlab application performed through artificial neural networks.

Keywords: