arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2403.01946cs.LG
收藏

一种对称变换的生成模型

A Generative Model of Symmetry Transformations

University of Cambridge(剑桥大学);MPI for Intelligent Systems, Tübingen(图宾根智能系统MPI(马克斯·普朗克智能系统研究所));Ångstrom AI;University of Amsterdam(阿姆斯特丹大学)

James Urquhart Allingham, Bruno Kacper Mlodozeniec, Shreyas Padhy, Javier Antorán, David Krueger, Richard E. Turner, Eric Nalisnick, José Miguel Hernández-Lobato

AI总结:

本文提出一种受群论启发的生成模型,可从数据中学习预指定对称变换的实际存在程度,并以可解释方式捕捉仿射与颜色对称性,从而提升测试对数似然和数据效率。

摘要:

正确捕捉数据的对称变换可以带来具有强泛化能力的高效模型,尽管融合对称性的方法通常需要先验知识。尽管近期在直接从数据集学习这些对称性方面已有进展,但这类工作大多聚焦于判别式场景。在本文中,我们受群论思想启发,构建了一个明确旨在捕捉数据近似对称性的生成模型。由此得到的模型在给定一组预先指定但范围广泛的可能对称性时,能够学习这些对称性实际存在的程度(如果存在的话)。我们的模型可被视为一种用于数据增强的生成过程。我们给出了学习该生成模型的简单算法,并通过实验证明其能够以可解释的方式捕捉仿射变换和颜色变换下的对称性。将我们的对称模型与标准生成模型相结合,可获得更高的边际测试对数似然,并提升数据效率。

英文摘要:

Correctly capturing the symmetry transformations of data can lead to efficient models with strong generalization capabilities, though methods incorporating symmetries often require prior knowledge. While recent advancements have been made in learning those symmetries directly from the dataset, most of this work has focused on the discriminative setting. In this paper, we take inspiration from group theoretic ideas to construct a generative model that explicitly aims to capture the data's approximate symmetries. This results in a model that, given a prespecified but broad set of possible symmetries, learns to what extent, if at all, those symmetries are actually present. Our model can be seen as a generative process for data augmentation. We provide a simple algorithm for learning our generative model and empirically demonstrate its ability to capture symmetries under affine and color transformations, in an interpretable way. Combining our symmetry model with standard generative models results in higher marginal test-log-likelihoods and improved data efficiency.

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。

↑