Valuating Cryptocurrency Assets using Linear Regression, HRL, and LSTM: Machine Learning Evidence

Since cryptocurrencies have grown in popularity as a form of investment, many researchers and investors are now interested in making predictions about their future value. This article seeks to explore and analyze three machine learning models for predicting bitcoin prices, i.e., linear regression, hierarchical reinforcement learning (HRL), and long short-term memory (LSTM). According to the findings, the Random Forest model fared better at predicting Bitcoin prices than other conventional machine learning models like Linear Regression and Support Vector Regression. With a low Mean Absolute Percentage Error, the HRL model, which is based on sentiment analysis of social media data, demonstrated encouraging results in predicting bitcoin prices. Last but not least, the deep learning-based LSTM model beat other models at predicting the price of bitcoin. These models have drawbacks because they are based on past data and might not take quick market shifts or unforeseen events into consideration. Future studies could investigate how real-time data and news stories can be used to increase the predictive power of machine learning models for cryptocurrencies. Overall, this work demonstrates the potential of machine learning in forecasting financial markets and adds to the expanding body of literature on cryptocurrency price prediction. The findings of this study may help traders and investors make wise selections in the bitcoin market.