BitForecast: CNN-LSTM Cryptocurrency Price Prediction Code

implementation code for the research paper "BitForecast: A Hybrid CNN-LSTM Model for Cryptocurrency Price Prediction". Includes: 1. Random Forest classifier for cryptocurrency price direction prediction (54.2% accuracy) 2. CNN-LSTM hybrid deep learning model (48.1% accuracy) 3. Feature engineering with technical indicators 4. Visualization code for generating paper figures 5. Requirements and setup instructions Tested on Bitcoin daily OHLCV data (January 2020 - July 2021). Dataset available on Kaggle: https://www.kaggle.com/datasets/sudalairajkumar/cryptocurrencypricehistory本代码为研究论文《BitForecast: 一种用于加密货币价格预测的混合CNN-LSTM模型》的实现代码，包含以下内容： 1. 用于加密货币价格走势预测的随机森林（Random Forest）分类器，分类准确率达54.2% 2. 混合卷积神经网络-长短期记忆网络（CNN-LSTM）深度学习模型，预测准确率达48.1% 3. 基于技术指标的特征工程实现 4. 用于生成论文配图的可视化代码 5. 环境依赖与配置安装说明 本代码已基于2020年1月至2021年7月的比特币每日OHLCV（开盘价、最高价、最低价、收盘价、成交量）数据完成测试。相关数据集可在Kaggle平台获取：https://www.kaggle.com/datasets/sudalairajkumar/cryptocurrencypricehistory