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 2102.10867cs.LGcs.AI
收藏

用于不变性发现的线性单元测试

Linear unit-tests for invariance discovery

Facebook AI Research(Facebook AI研究院);INRIA - PSL Research University(法国国家信息与自动化研究所 - 巴黎文理研究大学)

Benjamin Aubin, Agnieszka Słowik, Martin Arjovsky, Leon Bottou, David Lopez-Paz

摘要:

人们对学习跨训练环境的不变相关性的算法越来越感兴趣。目前的提议中有很大一部分在因果关系文献中找到了理论支持,但是,它们在实践中有多大用处?本说明的目的是提出六个线性低维问题--单元测试--以精确的方式评估不同类型的分布外泛化。经过初步实验,最近提出的三种替代方案都没有通过所有测试。通过提供代码来自动复制本文(https://www.github.com/facebookresearch/InvarianceUnitTests)中的所有结果,我们希望我们的单元测试成为分布外泛化研究人员的标准垫脚石。

英文摘要:

There is an increasing interest in algorithms to learn invariant correlations across training environments. A big share of the current proposals find theoretical support in the causality literature but, how useful are they in practice? The purpose of this note is to propose six linear low-dimensional problems -- unit tests -- to evaluate different types of out-of-distribution generalization in a precise manner. Following initial experiments, none of the three recently proposed alternatives passes all tests. By providing the code to automatically replicate all the results in this manuscript (https://www.github.com/facebookresearch/InvarianceUnitTests), we hope that our unit tests become a standard steppingstone for researchers in out-of-distribution generalization.

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。

↑