arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2209.00905cs.LGcond-mat.dis-nncond-mat.stat-mechphysics.chem-phphysics.comp-ph
收藏

从潜动力学到有意义的表示

From latent dynamics to meaningful representations

Institute for Physical Science and Technology, University of Maryland(马里兰大学物理科学与技术研究所);Department of Mathematics, University of Maryland(马里兰大学数学系)

Dedi Wang, Yihang Wang, Luke Evans, Pratyush Tiwary

AI总结:

本文提出纯动力学约束的表示学习框架,通过可学习转移密度的过阻尼 Langevin 动力学替代人工先验,并在多系统及荧光 DNA 数据上验证其可唯一识别有意义潜表示。

摘要:

尽管表示学习一直是机器学习和人工智能兴起的核心,但如何使学习到的表示具有意义仍是一个关键问题。为此,典型方法是通过先验概率分布对学习到的表示进行正则化。然而,这类先验通常不可获得,或者具有临时性。为解决这一问题,近期研究已转向利用物理原理所提供的洞见来指导学习过程。本着这一精神,我们提出了一个纯粹受动力学约束的表示学习框架。我们不依赖预定义概率,而是限制潜表示遵循具有可学习转移密度的过阻尼 Langevin 动力学——这是一种由统计力学驱动的先验。我们表明,对于随机动力系统中的表示学习而言,这是一种更自然的约束,并具有唯一识别真实表示的关键能力。我们在不同系统上验证了该框架,包括一个真实世界的荧光 DNA 电影数据集。结果表明,我们的算法能够唯一识别正交、等距且有意义的潜表示。

英文摘要:

While representation learning has been central to the rise of machine learning and artificial intelligence, a key problem remains in making the learned representations meaningful. For this, the typical approach is to regularize the learned representation through prior probability distributions. However, such priors are usually unavailable or are ad hoc. To deal with this, recent efforts have shifted towards leveraging the insights from physical principles to guide the learning process. In this spirit, we propose a purely dynamics-constrained representation learning framework. Instead of relying on predefined probabilities, we restrict the latent representation to follow overdamped Langevin dynamics with a learnable transition density - a prior driven by statistical mechanics. We show this is a more natural constraint for representation learning in stochastic dynamical systems, with the crucial ability to uniquely identify the ground truth representation. We validate our framework for different systems including a real-world fluorescent DNA movie dataset. We show that our algorithm can uniquely identify orthogonal, isometric and meaningful latent 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。

↑