Development of cryptocurrencies and analysis of the their place in the currency markets

In this study, the basic factors affecting cryptocurrency price movements, the applicability of short-term and long-term price predictions, and from which variables affected these predictions were investigated. In addition, the question of which artificial intelligence method should be preferred in order to obtain the most reliable estimate with minimum error was addressed. In this context, a model was developed to determine the closest prediction range to the truth using deep neural networks, and the findings obtained were intended to be a guide for investors and market makers. In the study, an approach that integrates different artificial intelligence techniques to improve price predictions in cryptocurrency markets was presented. Time series data were analyzed for stationarity wise and ADF, PP and KPSS tests have been applied for the purpose evaluate the impact on forecast accuracy. Realized predictions with the VAE-GAN model revealed that the model adapts better to volatile market conditions when trained with stabilized data, while analyses made with non-stabilized data provided higher accuracy rates for assets with high market value and trading volume. The findings show that AI-based modeling offers a powerful method for cryptocurrency price prediction. Keywords: Cryptocurrency, digital transformation, artificial intelligence, deep learning, VAE-GAN.