Scholarlink

Cryptocurrency, particularly Bitcoin, has become a significant component of the financial asset market over the last years with its unstable price. For Traders, investors, and other market participants, the potential movement in Bitcoin’s price is becoming more important. This study focuses on mitigating the unpredictability of Bitcoin prices by using the classifier of XGBoost. This work uses historical data of Bitcoin’s price and data of relevant factors like trading volume, sentiment analysis of the market, and technical indicator. Data pre-processing techniques such as missing value treatment, normalization, and the split of data into training and testing are performed first. XGBoost is then trained in the developing set to identify the patterns in the input variables. The test data is then employed to assess the model with the help of various metrics including accuracy, precision, recall, and F1-score. The XGBoost model has also been compared with the other base models and the traditional forecasting models based on time series. Our results also suggest that the proposed model outperformed the benchmark model and reported 97% accuracy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Bitcoin Price Prediction Using XGBoost Classifier: A Machine Learning Approach

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • P. H. V. Sesha Talpa Sai,
  • Ganesh Yellappa Acharya,
  • S. Pramod,
  • Shanmuga Canaraj Ganesan,
  • Saripelli Chaitanya Bangarraju,
  • Kishan Tiwari,
  • Katakam Jaswanthi,
  • Amiya Bhaumik

摘要

Cryptocurrency, particularly Bitcoin, has become a significant component of the financial asset market over the last years with its unstable price. For Traders, investors, and other market participants, the potential movement in Bitcoin’s price is becoming more important. This study focuses on mitigating the unpredictability of Bitcoin prices by using the classifier of XGBoost. This work uses historical data of Bitcoin’s price and data of relevant factors like trading volume, sentiment analysis of the market, and technical indicator. Data pre-processing techniques such as missing value treatment, normalization, and the split of data into training and testing are performed first. XGBoost is then trained in the developing set to identify the patterns in the input variables. The test data is then employed to assess the model with the help of various metrics including accuracy, precision, recall, and F1-score. The XGBoost model has also been compared with the other base models and the traditional forecasting models based on time series. Our results also suggest that the proposed model outperformed the benchmark model and reported 97% accuracy.