LOB-Bench
收藏资源简介:
LOB-Bench是由牛津大学等机构开发的一个金融领域生成式AI模型评估基准。该数据集包含按照LOBSTER格式组织的限制订单簿数据,旨在评估生成式模型在金融时间序列数据上的表现。数据集涵盖了Alphabet Inc (GOOG)和Intel Corporation (INTC)股票的数据,通过比较生成数据和真实数据在多个维度上的分布差异,为模型性能提供了全面的量化评估。这一基准不仅易于使用和访问,还可以扩展到其他高维时间序列任务领域。
LOB-Bench is a financial domain generative AI model evaluation benchmark developed by institutions such as the University of Oxford. This dataset contains limit order book data organized in the LOBSTER format, designed to evaluate the performance of generative models on financial time series data. The dataset covers data for stocks of Alphabet Inc. (GOOG) and Intel Corporation (INTC), and provides a comprehensive quantitative assessment of model performance by comparing the distribution differences between generated data and real data across multiple dimensions. This benchmark is not only easy to use and access, but also extendable to other high-dimensional time series task domains.




