Abstract:

Cryptocurrencies are peer-to-peer transaction systems that use the secure hash algorithm (SHA)-256 and message digest (MD)- 5 methods to protect data transactions. Cryptocurrency values are exceedingly volatile, follow stochastic moments, and have achieved unpredictability. They are frequently used for investment and have replaced traditional forms of investment like as metals, estates, and the stock market. Their commercial prominence necessitates the creation of a strong forecasting model. However, given to its reliance on other cryptocurrencies, bitcoin price forecast is difficult. Many academics have employed machine learning and deep learning models, as well as other market sentiment-based algorithms, to forecast cryptocurrency prices. Because all cryptocurrencies belong to the same class, a rise in the price of one cryptocurrency might cause a price change for other cryptocurrencies. The emotions from tweets and other social media platforms were also used by the researchers to improve the performance of their suggested system. Motivated by this, we offer in this study a hybrid and resilient framework, DL-Gues, for cryptocurrency price prediction that takes into account its interdependence on other cryptocurrencies as well as market attitudes. For validation, we investigated Dash price prediction utilising price history and tweets of Dash, Litecoin, and Bitcoin for different loss functions. To test the applicability of DL-GuesS on additional cryptocurrencies, we inferred findings for Bitcoin-Cash price prediction using the price history and tweets of Bitcoin-Cash, Litecoin, and Bitcoin.

Description:

Cryptocurrency, complex systems, fusion of cryptocurrency, price prediction, VADER, sentiment analysis, deep learning, systems of systems.