arXiv：cs.LG（机器学习，全量分类）· Simon Forbat, Rainer Gemulla·· 14 小时前AI 评分35
RelICL：用表格基础模型实现免训练的关系学习
RelICL: Training-free Relational Learning with Tabular Foundation Models
AI 导读
RelICL 是一种免训练的关系学习方法，通过沿 schema 图逐步传播并融合信息，缓解深度特征合成（DFS）的特征爆炸与交互盲区两大问题，且复用最终用于预测的同一表格基础模型。在 RelBench 任务上，RelICL 与最强的 DFS 方法表现相当。
正文
Abstract:Tabular foundation models achieve state-of-the-art performance on single-table tasks without any training. Recent work suggests that they are also well-suited for relational learning via deep feature synthesis (DFS), which flattens a relational schema into a single table by adding aggregates of the other tables' columns as features. This approach is appealing because it directly benefits from improvements to or customization of the underlying tabular foundation model. In this paper, we identify two key problems with DFS: feature explosion and interaction blindness. The first problem arises because the number of DFS features grows quickly as the schema becomes more complex, limiting scalability and performance. The second problem arises because column-wise aggregates do not account for feature interactions, limiting performance. We propose and explore an alternative method termed RelICL, which keeps the benefits of DFS but alleviates these two problems. At its heart, RelICL propagates and fuses information step by step through the schema graph, using the same tabular foundation model that is eventually used for prediction to do so. In our experimental study using RelBench tasks, RelICL was on par with the strongest approach based on deep feature synthesis.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.01725 [cs.LG] |
| (or arXiv:2610.01725v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.01725 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Simon Forbat [view email]
[v1]
Thu, 1 Oct 2026 14:00:23 UTC (67 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org