PMLB v1.0
收藏资源简介:
PMLB v1.0是由宾夕法尼亚大学开发的一个大型开源数据集集合,旨在为机器学习和数据科学方法的评估提供标准化比较。该数据集集合整合了来自UCI ML仓库和OpenML等多个来源的数百个公开可用数据集,支持回归和分类问题。数据集创建过程中,每个数据集都配有详细的元数据描述和自动生成的Pandas Profiling报告,以确保数据质量。PMLB的应用领域广泛,主要用于评估和比较新的机器学习算法,解决数据集访问困难、集成分析挑战和数据质量问题。
PMLB v1.0 is a large-scale open-source dataset collection developed by the University of Pennsylvania, designed to provide standardized comparative benchmarks for evaluating machine learning and data science methodologies. This collection integrates hundreds of publicly available datasets from multiple sources including the UCI Machine Learning Repository and OpenML, supporting both regression and classification tasks. During the dataset curation process, each dataset is accompanied by detailed metadata descriptions and automatically generated Pandas Profiling reports to ensure data quality. Widely applicable across diverse domains, PMLB is primarily employed to evaluate and compare emerging machine learning algorithms, while also addressing core challenges including limited dataset accessibility, barriers to integrated analysis, and data quality concerns.




