Cryptocurrency Price Prediction using Regression Models on Momentum Indicators

Cryptocurrencies have enormously gained popularity ever since the launch of Bitcoin in 2009 and are now valued at over a trillion dollars as a whole. A significant number of people actively participate in cryptocurrency trading, to the extent that various goods and services can be acquired through the use of digital currencies like Bitcoin, Ethereum, and others. Nevertheless, the analysis of trends, graphs, and trading strategies involves a considerable level of complexity. Making use of conducted research enables traders to effectively capitalize on trading opportunities. The proposed model uses three momentum indicators which are the Relative Strength Index(RSI), Moving Average Convergence Divergence(MACD), and Average Directional Index(ADX) as features. Each indicator takes as input custom factors that give substantial trend analysis. The values of each indicator are used as features in a variety of machine-learning models that rely on regression to predict the price of cryptocurrencies in the future. The price prediction model is fairly accurate, and it is observed that the gradient boost shows the highest accuracy of the three models with an R-squared value equal to 0.968.