Short Term and Current Bitcoin Price Prediction with Machine Lear-ning Methods

In today's world, where change or innovation in one field rapidly affects all fields, technological developments provide both recording, guiding and facilitating this interaction process. With the development of the Internet, the money used as cash has been digitalised and has become the numbers in bank accounts. Money has carried many meanings and features for humanity and societies since its emergence. It has been an irreplaceable invention for the economy and finance sector with its many functions. In 2009, with the introduction of Bitcoin, the concepts of virtual money and cryptocurrency were added to the variety of money. This development of money has increased the emergence and importance of cryptocurrencies. Cryptocurrency is a combination of the words crypto and currency. The security of cryptocurrencies is ensured by using cryptography protected by mathematical ciphers. In addition, cryptocurrencies have become an exchange and investment tool that is not under the control of a central authority and has rapidly gained popularity in the financial sectors. Cryptocurrencies have many variables that affect the exchange volume of cryptocurrencies, making it more difficult to predict cryptocurrencies than other investment instruments. Therefore, it is of great importance for investors to develop intelligent forecasting models in order to make the right investment decisions. Bitcoin is known as the world's most valuable cryptocurrency and is traded on many exchanges. A high amount of transactions occur on a daily basis. In this case, traditional forecasting methods may not be sufficient for Bitcoin price prediction. In this study, artificial intelligence-based deep learning models such as Long iv Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) are used for short-term price prediction with current price data of Bitcoin. The data set was obtained from Yahoo's public API using the y-finance library. In addition, a literature study on machine learning, deep learning and performance metrics on blockchain technology and Bitcoin was conducted. 2023, ix + 44 pages Keywords: Bitcoin, Deep Learning, Price Forecast, GRU, LSTM