sovai/price_breakout

该数据集提供了对美国股票价格向上突破的每日预测，涵盖超过13,000只美国股票。预测的准确率约为65%，ROC-AUC为68%。数据集每周重新训练，使用了多种机器学习模型，包括校准分类器、专有回归模型和符合回归模型。数据集还提供了预测的概率分数和置信区间。数据集的访问方式包括获取最新数据、完整历史数据和特定股票的数据。数据字典详细列出了数据集的各个列及其描述。

This dataset identifies potential price breakout stocks over the next 30-60 days for US Equities. It provides daily predictions of upward price breakouts for over 13,000 US equities. The dataset has an accuracy of around 65% and an ROC-AUC of 68%, making it one of the most accurate breakout models on the market. The dataset is retrained on a weekly basis and uses several machine learning models, including classification algorithms, regression models, and conformal predictors, to generate price movement predictions, probability scores, and confidence intervals. The dataset also includes standard deviations and lower and upper bounds of confidence intervals to assess the uncertainty of the predictions. Data access methods include retrieving the latest data, full history, or specific tickers. Additionally, the data dictionary details the columns and their descriptions, as well as use cases for professional investors such as portfolio optimization, risk management, sector and market analysis, and market timing.