Abstract:
A cryptocurrency is a digital currency based on a decentralized blockchain network. Transactions
in cryptocurrencies are performed without a central authority or single administration.
The transactions are secured by the strong hashing algorithm (SHA-256). There are
more than 22,000 cryptocurrencies in the market, with over a $1 Trillion market cap. BTC
(Bitcoin) is a famous cryptocurrency, designed to be a decentralized and secure form of
digital cash. Cryptocurrencies can be used to purchase goods and services, transfer funds,
and even as investments. The use of cryptocurrency has increased in the last few years,
cryptocurrencies are commonly used for investment purposes.
This research focuses on developing a price prediction model for Dash coin and Bitcoin
Cash. The prices of cryptocurrencies are highly unstable, which makes it challenging to
forecast future prices. Researchers used Twitter sentiments, news, and previous market data
with the help of NLP, Machine Learning (ML), and Deep Learning (DL) techniques to predict
the future prices of different currencies. Our research uses state-of-the-art DL techniques
to build a prediction model. The inclusion of technical indicators such as the Relative
Strength Index and Moving Average along with the Fear & Greed Index and historic data
helps to capture market sentiment and improve the overall accuracy of the model. We trained
Gradient Recurrent Unit (GRU) model for Bitcoin Cash, and Dash cryptocurrencies, and the
results show that our approach has better results than others.
Our work focuses on predicting future close prices of cryptocurrencies. Our main objective
is to incorporate additional key features such as technical indicators and the Fear & Greed
Index data can lead to improved accuracy for existing market models available. This research
will help investors to better understand the direction in which the market is moving.
