Bitcoin Price Prediction Using Machine Learning Algorithm

Abstract

Bitcoin, being one of the top cryptocurrencies, sees its price influenced by market trends and public opinions. This study introduces a prediction model that uses a Random Forest Classifier along with sentiment analysis to forecast Bitcoin prices. Data on Bitcoin’s past and current performance was gathered using the finance library, while sentiment analysis was conducted on news and comments related to Bitcoin, collected through mwclient. The model’s success was measured using accuracy, precision, recall, and F1-score. The findings show that combining market data with sentiment analysis helps in making more accurate predictions. The model leverages a Random Forest Classifier, which is trained on historical Bitcoin price data sourced from the finance library. In addition, sentiment analysis is conducted on news articles and public discussions related to Bitcoin, collected using the mwclient library, to assess the prevailing public sentiment and its potential influence on price fluctuations. The model’s performance is measured using key evaluation metrics such as accuracy, precision, recall, and F1-score, offering a thorough assessment of its effectiveness. The findings reveal that incorporating sentiment analysis alongside market data significantly boosts the model's predictive power when compared to traditional models that rely solely on past price data. This integrated approach holds promise for enhancing price prediction models in the cryptocurrency market, where factors like public sentiment and social dynamics play a crucial role in shaping price movements.