arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.00482cs.LGcs.AIcs.CVcs.CYcs.ITmath.IT
收藏

利用部分信息分解形式化有偏数据集的虚假性

Towards Formalizing Spuriousness of Biased Datasets Using Partial Information Decomposition

University of Maryland, College Park(马里兰大学学院公园分校);Google Research(谷歌研究);Princeton University(普林斯顿大学)

Barproda Halder, Faisal Hamman, Pasan Dissanayake, Qiuyi Zhang, Ilia Sucholutsky, Sanghamitra Dutta

AI总结:

针对有偏数据集的虚假关联问题,提出基于PID的Spurious Disentangler框架,在训练前拆解信息分量、度量数据集虚假性,经6个基准数据集验证其与训练后模型泛化指标一致。

摘要:

虚假性产生于数据集中两个或多个不存在因果关系的变量之间存在关联的情况。本研究提出一个可解释性框架,在模型训练前预先拆解数据集中此类虚假关联的本质。我们利用信息论中名为部分信息分解(Partial Information Decomposition,PID)的研究成果,将关于目标的全部信息分解为四个非负分量,即唯一信息(分别存在于核心特征与虚假特征中)、冗余信息和协同信息。该框架可帮助预判:在基于该数据集训练最优分类器时,核心特征或虚假特征何时不可或缺、二者何时可单独满足需求、何时需要二者共同作用。接下来,我们基于该分解提出一种新的数据集虚假性度量方法。我们通过检验若干候选度量方法,借助直观的典型示例与反例展示其捕捉到的内容与遗漏的信息,系统性地推导出该度量。我们提出的Spurious Disentangler框架包含分割、降维与估计模块,具备高效处理高维图像数据的能力。最后,我们开展实证评估,在6个基准数据集的多种实验设置下,展示唯一信息、冗余信息、协同信息的变化趋势,以及我们提出的虚假性度量的表现。我们观察到,所提出的数据集虚假性预度量与最差组准确率等训练后模型泛化指标具有一致性,进一步支撑了我们的结论。代码可在https://github.com/Barproda/spuriousness-disentangler获取。

英文摘要:

Spuriousness arises when there is an association between two or more variables in a dataset that are not causally related. In this work, we propose an explainability framework to preemptively disentangle the nature of such spurious associations in a dataset before model training. We leverage a body of work in information theory called Partial Information Decomposition (PID) to decompose the total information about the target into four non-negative quantities, namely unique information (in core and spurious features, respectively), redundant information, and synergistic information. Our framework helps anticipate when the core or spurious feature is indispensable, when either suffices, and when both are jointly needed for an optimal classifier trained on the dataset. Next, we leverage this decomposition to propose a novel measure of the spuriousness of a dataset. We arrive at this measure systematically by examining several candidate measures, and demonstrating what they capture and miss through intuitive canonical examples and counterexamples. Our framework Spurious Disentangler consists of segmentation, dimensionality reduction, and estimation modules, with capabilities to specifically handle high-dimensional image data efficiently. Finally, we also perform empirical evaluation to demonstrate the trends of unique, redundant, and synergistic information, as well as our proposed spuriousness measure across $6$ benchmark datasets under various experimental settings. We observe an agreement between our preemptive measure of dataset spuriousness and post-training model generalization metrics such as worst-group accuracy, further supporting our proposition. The code is available at https://github.com/Barproda/spuriousness-disentangler.

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。

↑