Hybrid Data Decomposition-based Deep Learning for Bitcoin Prediction and Algorithm Trading

Abstract

Since the beginning of time, money has played a crucial role in both storage and the exchange of goods and services.
Fiat money, which is issued and managed by centralized banks and governments, is the modern form of money. Fiat
money has several benefits, such as increased affordability, growth potential, economic stability, and global use, yet
inflation results from printing more of it. Bitcoin operates on a highly secure distributed network's bottom root.
Bitcoin is used for both transactions and asset storage because of its global acceptability. Bitcoin's primary benefits
include being impenetrable and preventing double spending after a transaction is completed. The Bitcoin transaction
is entirely transparent to all users on the network. As a result of its increased public appeal, Bitcoin continues to hold
the top spot. The Deep Learning model of Multiplicative LSTM Networks is coupled with the Attention Mechanism
for Bitcoin's next price prediction utilizing Technical Indicators in order to circumvent the gradient descent issue.
Pattern-based indicators known as technical indicators are used to uncover hidden information in Bitcoin's historical
raw data.