arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.10441cs.LGcs.AI
收藏

LakeMLB: 数据湖机器学习基准

LakeMLB: Data Lake Machine Learning Benchmark

Shanghai Jiao Tong University(上海交通大学);Beijing Institute of Technology(北京理工大学);Nankai University(南开大学)

Feiyu Pan, Tianbin Zhang, Aoqian Zhang, Yu Sun, Zheng Wang, Lixing Chen, Li Pan, Jianhua Li

AI总结:

LakeMLB提出一个数据湖机器学习基准,针对多源多表场景,提供真实数据集和三种整合策略,评估机器学习方法在复杂数据湖环境中的性能。

摘要:

现代数据湖已成为大规模机器学习的基础平台,通过表导向的抽象方法,使异构数据的灵活存储和结构化分析成为可能。尽管其重要性日益增长,但评估数据湖环境中机器学习性能的标准基准仍然稀缺。为解决这一缺口,我们提出了LakeMLB(数据湖机器学习基准),旨在最常见多源、多表场景下的数据湖环境。LakeMLB专注于两种具有代表性的多表场景,即Union和Join,并为每个场景提供三个真实世界的数据集,涵盖政府开放数据、金融、维基百科和在线市场。该基准支持三种具有代表性的整合策略:基于预训练的方法、基于数据增强的方法以及基于特征增强的方法。我们使用最先进的表格学习方法进行了广泛的实验,提供了在复杂数据湖场景下这些方法性能的洞察。我们发布了数据集和代码,以促进对数据湖生态系统中机器学习的严格研究;该基准可在https://github.com/zhengwang100/LakeMLB上获得。

英文摘要:

Data lakes have become a fundamental platform for large-scale machine learning by enabling flexible management of heterogeneous data. Despite their growing importance, standardized benchmarks for evaluating machine learning performance in data lake environments remain scarce. To address this gap, we present LakeMLB (Data Lake Machine Learning Benchmark), the first benchmark designed for multi-table machine learning in data lakes. LakeMLB focuses on two representative scenarios, Union and Join, and provides six real-world datasets spanning diverse domains. It supports three representative multi-table learning paradigms: pre-training, data augmentation, and feature augmentation, together with standardized data splits and evaluation protocols. We conduct extensive experiments with state-of-the-art tabular learning methods and provide insights into their performance across different data lake scenarios. We release both datasets and code to facilitate rigorous research on machine learning in data lake ecosystems; the benchmark is available at https://github.com/zhengwang100/LakeMLB.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑