arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2209.14734cs.LG
收藏

DiGress:用于图生成的离散去噪扩散模型

DiGress: Discrete Denoising diffusion for graph generation

EPFL(洛桑联邦理工学院)

Clement Vignac, Igor Krawczuk, Antoine Siraudin, Bohan Wang, Volkan Cevher, Pascal Frossard

AI总结:

本文提出 DiGress,一种用于图生成的离散去噪扩散模型,通过图 Transformer 逆转加噪过程并引入马尔可夫噪声模型及辅助特征,在分子与非分子数据集上达到 SOTA,且首次无需分子特定表示即扩展至大型 GuacaMol 数据集。

摘要:

本工作提出了 DiGress,一种用于生成具有类别型节点和边属性的图的离散去噪扩散模型。该模型利用离散扩散过程,通过添加或删除边以及改变类别,逐步向图中注入噪声。通过训练图 Transformer 网络来逆转这一过程,将图上的分布学习问题简化为一系列节点和边分类任务。我们进一步通过引入在扩散过程中保持节点和边类型边缘分布的马尔可夫噪声模型,并结合辅助图论特征,提升了样本质量。此外,还提出了一种以图级特征为条件进行生成的方法。DiGress 在分子和非分子数据集上均取得了最先进的性能,在平面图数据集上的有效性提升高达 3 倍。它也是首个能够扩展至包含 130 万类药分子的 GuacaMol 大型数据集,且无需使用分子特定表示的模型。

英文摘要:

This work introduces DiGress, a discrete denoising diffusion model for generating graphs with categorical node and edge attributes. Our model utilizes a discrete diffusion process that progressively edits graphs with noise, through the process of adding or removing edges and changing the categories. A graph transformer network is trained to revert this process, simplifying the problem of distribution learning over graphs into a sequence of node and edge classification tasks. We further improve sample quality by introducing a Markovian noise model that preserves the marginal distribution of node and edge types during diffusion, and by incorporating auxiliary graph-theoretic features. A procedure for conditioning the generation on graph-level features is also proposed. DiGress achieves state-of-the-art performance on molecular and non-molecular datasets, with up to 3x validity improvement on a planar graph dataset. It is also the first model to scale to the large GuacaMol dataset containing 1.3M drug-like molecules without the use of molecule-specific representations.

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。

↑