Estimation of the value of crypto currency using neural networks

Today, people make purchases and sales transactions in different exchanges in order to increase their income. With the development of technology, cryptocurrency exchanges have become one of the exchanges that people use to earn more income. Technical and fundamental analysis methods are used while trading in stock markets. Technical analysis is the process of predicting future price movements based on historical data. When performing technical analysis, very large data can be encountered. For this reason, the analysis of the data becomes difficult and the probability of the data to be obtained as a result of technical analysis increases. As a result of this situation, investors who cannot analyze big data correctly are more likely to face big losses. Due to the emergence of cryptocurrencies in the near future, there are not many studies in the literature on predicting the price of cryptocurrencies. Cryptocurrency prediction is valuable both for investors to make the right decision and because it is open to applications in the scientific field. For this reason, in this study, estimation was carried out by choosing the 3 cryptocurrencies with the highest cryptocurrency mobility. Selected cryptocurrencies; Bitcoin, Ethereum and Cardano. Due to the large size of the data and in terms of the analysis of decision factors, the closing value of the next day was estimated by using the opening, closing, smallest and largest values of these cryptocurrencies with Artificial Neural Networks and Regression Analysis methods. Afterwards, a comparison was made between the estimated values and the actual values. As a result of the study, it was observed that the estimation study made with Artificial Neural Networks performed more successfully than the estimation result made with Regression Analysis.