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 2602.07875cs.LG
收藏

Harpoon:基于条件表格扩散的通用流形引导

Harpoon: Generalised Manifold Guidance for Conditional Tabular Diffusion

Department of Computer Science, Delft University of Technology(代尔夫特理工大学计算机科学系);Department of Computer Science, Université de Neuchâtel(日内瓦大学计算机科学系)

Aditya Shankar, Yuandou Wang, Rihan Hai, Lydia Y. Chen

AI总结:

本文提出Harpoon,一种基于流形引导的条件表格扩散方法,通过扩展流形理论来处理多样化的推理目标,从而在表格数据生成中实现更精确的条件控制。

摘要:

在需要对生成过程进行精确控制的应用中,生成表格数据至关重要。现有方法依赖于训练时的策略,无法在推理时泛化到未见过的约束,并且难以处理超出表格填补的条件任务。虽然流形理论提供了一种指导生成的原理化方法,但当前的公式化方法局限于特定的推理时间目标,并且仅限于连续领域。我们扩展了流形理论到表格数据,并扩展了其范围以处理多样的推理时间目标。在此基础上,我们引入了HARPOON,一种表格扩散方法,通过引导无约束样本沿着流形几何来满足多样化的表格条件。我们在填补和强制不等约束等任务上通过实验证明了我们的理论贡献,展示了HARPOON在各种数据集上的强大性能以及流形感知指导在表格数据中的实际好处。代码URL:https://github.com/adis98/Harpoon

英文摘要:

Generating tabular data under conditions is critical to applications requiring precise control over the generative process. Existing methods rely on training-time strategies that do not generalise to unseen constraints during inference, and struggle to handle conditional tasks beyond tabular imputation. While manifold theory offers a principled way to guide generation, current formulations are tied to specific inference-time objectives and are limited to continuous domains. We extend manifold theory to tabular data and expand its scope to handle diverse inference-time objectives. On this foundation, we introduce HARPOON, a tabular diffusion method that guides unconstrained samples along the manifold geometry to satisfy diverse tabular conditions at inference. We validate our theoretical contributions empirically on tasks such as imputation and enforcing inequality constraints, demonstrating HARPOON'S strong performance across diverse datasets and the practical benefits of manifold-aware guidance for tabular data. Code URL: https://github.com/adis98/Harpoon

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。

↑