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 2605.30153stat.MLcs.ITcs.LGmath.ITmath.STstat.TH
收藏

扩散模型在学习低维多模态分布时具有统计最优性

Diffusion Models Are Statistically Optimal for Learning Low-Dimensional Multi-Modal Distributions

Department of Industrial and Operations Engineering, University of Michigan, Ann Arbor, USA(工业与运营工程系,密歇根大学,安娜堡,美国)

Jingda Wu, Changxiao Cai

AI总结:

本文证明扩散模型在学习支撑在低维子空间并集上的分布时,样本复杂度仅依赖于内在维度,达到近最优的1-Wasserstein误差率,无需光滑性或有界密度假设。

摘要:

基于分数的扩散模型在学习高维分布,特别是那些具有低维和多模态结构的分布方面,已经展现出显著的实证成功。然而,对其统计效率的理论理解仍然有限。现有理论通常依赖于强正则性假设,例如一致有界密度或全局光滑的分数函数,这些假设无法捕捉此类内在结构。在这项工作中,我们研究了扩散模型在学习支撑在低维子空间并集上的分布时的样本复杂度。假设每个子空间内的数据分布是次高斯的,我们证明扩散模型最多需要$\widetilde{O}(\varepsilon^{-k \vee 2})$个样本即可在1-Wasserstein距离上达到$\varepsilon$误差,其中$k$是内在维度。这一近最优的收敛速率仅依赖于内在维度,并显著改进了先前遭受维度灾难的理论保证。值得注意的是,我们的分析适用于广泛的分布,无需施加光滑性、有界密度或对数凹性假设。总体而言,我们的结果表明,扩散模型能够统计适应内在低维结构,同时自然容纳多模态数据,为其在复杂高维学习任务中的成功提供了严格的理论依据。

英文摘要:

Score-based diffusion models have demonstrated remarkable empirical success in learning high-dimensional distributions, particularly those exhibiting low-dimensional and multi-modal structures. However, theoretical understanding of their statistical efficiency remains limited. Existing theories typically rely on strong regularity assumptions, such as uniformly bounded densities or globally smooth score functions, which fail to capture such intrinsic structures. In this work, we study the sample complexity of diffusion models for learning distributions supported on a union of low-dimensional subspaces. Assuming that the data distribution within each subspace is subgaussian, we show that diffusion models require at most $\widetilde{O}(\varepsilon^{-k \vee 2})$ samples to achieve $\varepsilon$ error in 1-Wasserstein distance, where $k$ is the intrinsic dimension. This near-optimal convergence rate depends only on the intrinsic dimension and significantly improves upon prior theoretical guarantees that suffer from the curse of dimensionality. Notably, our analysis applies to a broad collection of distributions without imposing smoothness, bounded-density, or log-concavity assumptions. Overall, our results show that diffusion models can statistically adapt to intrinsic low-dimensional structure while naturally accommodating multi-modal data, offering a rigorous theoretical justification for their success in complex high-dimensional learning tasks.

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。

↑