arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.00332cs.LGcs.CV
收藏

IID表示学习及其在生物医学数据上的应用

Towards IID representation learning and its application on biomedical data

University Hospital, University of Zurich(苏黎世大学附属医院);University of Bern(伯尔尼大学);City University of Hong Kong(香港城市大学);University of Oxford(牛津大学)

Jiqing Wu, Inti Zlobec, Maxime Lafarge, Yukun He, Viktor H. Koelzer

摘要:

由于现实世界数据的异质性,广泛接受的独立同分布(IID)假设在最近的因果关系研究中受到了批评。在本文中,我们认为,而不是一个有问题的假设,IID是一个基本的任务相关的属性,需要学习。考虑$k$独立随机向量$\mathsf{X}^{i = 1, \ldots, k}$,我们详细说明了如何可以重新制定各种不同的因果问题,以学习任务相关的功能$ϕ$,诱导IID之间的$\mathsf{Z}^i := ϕ\circ \mathsf{X}^i$,我们称之为IID表示学习。为了证明概念,我们研究了在分布外(OOD)泛化任务上的IID表示学习。具体而言,通过利用通过诱导IID的学习函数获得的表示,我们对两个生物医学数据集进行分子特征预测(分子预测),其中真实世界的分布偏移由a)分析前变异和b)采样协议引入。为了实现再现性并与最先进(SOTA)方法进行比较,请遵循WILDS推荐的OOD基准测试指南。与WILDS中支持的SOTA基线相比,结果证实了IID表示学习在OOD任务上的优越性能。该代码可通过https://github.com/CTPLab/IID_representation_learning公开访问。

英文摘要:

Due to the heterogeneity of real-world data, the widely accepted independent and identically distributed (IID) assumption has been criticized in recent studies on causality. In this paper, we argue that instead of being a questionable assumption, IID is a fundamental task-relevant property that needs to be learned. Consider $k$ independent random vectors $\mathsf{X}^{i = 1, \ldots, k}$, we elaborate on how a variety of different causal questions can be reformulated to learning a task-relevant function $ϕ$ that induces IID among $\mathsf{Z}^i := ϕ\circ \mathsf{X}^i$, which we term IID representation learning. For proof of concept, we examine the IID representation learning on Out-of-Distribution (OOD) generalization tasks. Concretely, by utilizing the representation obtained via the learned function that induces IID, we conduct prediction of molecular characteristics (molecular prediction) on two biomedical datasets with real-world distribution shifts introduced by a) preanalytical variation and b) sampling protocol. To enable reproducibility and for comparison to the state-of-the-art (SOTA) methods, this is done by following the OOD benchmarking guidelines recommended from WILDS. Compared to the SOTA baselines supported in WILDS, the results confirm the superior performance of IID representation learning on OOD tasks. The code is publicly accessible via https://github.com/CTPLab/IID_representation_learning.

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。

↑