tripolskypetr/btcusdt_feb_2026_candles_news_forecast
收藏资源简介:
BTCUSDT多时间框架GARCH波动率数据集是一个包含2026年2月(回测模拟)期间BTCUSDT交易对的GARCH预测波动率(sigma)的数据集。数据集以1分钟分辨率记录,共约40,320条记录。每条记录是8个不同时间框架下的GARCH预测波动率快照,并包含一个reliable标志,指示模型是否有足够的历史蜡烛图数据来产生统计稳定的估计。数据集以JSONL格式存储,适用于训练波动率预测模型、研究跨时间框架GARCH sigma相关性、基准波动率估计器以及为加密货币交易ML管道进行特征工程。
The BTCUSDT Multi-Timeframe GARCH Volatility Dataset contains GARCH-predicted volatility (sigma) for BTCUSDT across 8 timeframes during a backtest simulation in February 2026. The dataset records ~40,320 entries at 1-minute resolution. Each row is a 1-minute snapshot of GARCH-predicted volatility with a reliable flag indicating whether the model had enough historical candles to produce a statistically stable estimate. The dataset is stored in JSONL format and is suitable for training volatility prediction models, studying cross-timeframe GARCH sigma correlations, benchmarking volatility estimators, and feature engineering for crypto trading ML pipelines.




