arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.01136cs.LGcs.AI
收藏

TabINR:一种用于表格数据插补的隐式神经表示框架

TabINR: An Implicit Neural Representation Framework for Tabular Data Imputation

Department of Biomedical Engineering, Faculty of Medicine, University of Basel(巴塞尔大学生物医学工程系);Clarunis, Department of Visceral Surgery, Basel(Clarunis visceral外科部);Department of Visceral Surgery, Cantonal Hospital Basel-Land(巴塞尔-兰州 cantonal 医院 visceral 外科部);Department of Surgery, East Carolina University, Brody School of Medicine(东卡罗来纳大学外科系,Brody医学院)

Vincent Ochs, Florentin Bieder, Sidaty el Hadramy, Paul Friedrich, Stephanie Taha-Mehlitz, Anas Taha, Philippe C. Cattin

AI总结:

针对表格数据缺失导致下游模型性能下降的问题,提出基于自解码器的隐式神经表示框架TabINR,通过可学习行与特征嵌入实现自适应插补,在12个真实数据集上表现优于多数现有方法。

摘要:

表格数据是众多应用的基础,但现实世界的数据集常因采集错误、隐私限制或传感器故障而存在缺失。由于缺失值会降低下游模型的性能或阻碍其适用性,而简单的插补策略往往会引入偏差或扭曲底层数据分布,因此我们需要能够提供高质量插补结果、在不同规模数据集上均表现鲁棒且推理速度快的插补器。为此我们提出TabINR,这是一种基于自解码器的隐式神经表示(INR)框架,将表格建模为神经函数。基于可泛化INR的最新进展,我们引入了可学习的行嵌入和特征嵌入,能够有效处理表格数据的离散结构,且可从部分观测中推断,无需修改已训练模型即可实现实例自适应插补。我们在12个不同的真实世界数据集和多种缺失机制下评估了该框架,结果表明其插补精度始终表现优异,大多能匹配或优于经典模型(KNN、MICE、MissForest)和基于深度学习的模型(GAIN、ReMasker),在高维数据集上的提升最为显著。

英文摘要:

Tabular data builds the basis for a wide range of applications, yet real-world datasets are frequently incomplete due to collection errors, privacy restrictions, or sensor failures. As missing values degrade the performance or hinder the applicability of downstream models, and while simple imputing strategies tend to introduce bias or distort the underlying data distribution, we require imputers that provide high-quality imputations, are robust across dataset sizes and yield fast inference. We therefore introduce TabINR, an auto-decoder based Implicit Neural Representation (INR) framework that models tables as neural functions. Building on recent advances in generalizable INRs, we introduce learnable row and feature embeddings that effectively deal with the discrete structure of tabular data and can be inferred from partial observations, enabling instance adaptive imputations without modifying the trained model. We evaluate our framework across a diverse range of twelve real-world datasets and multiple missingness mechanisms, demonstrating consistently strong imputation accuracy, mostly matching or outperforming classical (KNN, MICE, MissForest) and deep learning based models (GAIN, ReMasker), with the clearest gains on high-dimensional datasets.

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。

↑