arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2002.03035math.OCstat.ML
收藏

Wasserstein近似梯度算法

The Wasserstein Proximal Gradient Algorithm

Adil Salim, Anna Korba, Giulia Luise

摘要:

Wasserstein梯度流是定义最速下降曲线以最小化概率测度空间上的目标函数的连续时间动态(即,Wasserstein空间)。这一目标通常是一个分歧w.r.t.固定的目标分布。近年来,这些连续时间动力学已被用于研究旨在近似概率分布的机器学习算法的收敛性。然而,这些算法的离散时间行为可能不同于连续时间动态。此外,虽然离散梯度流已经在文献中提出,很少知道他们的最小化能力。在这项工作中,我们提出了一个向前向后(FB)离散计划,可以解决的情况下,目标函数是一个光滑和非光滑测地线凸项的总和。使用技术从凸优化和最优输运,我们分析了FB计划作为一个最小化算法的Wasserstein空间。更确切地说,我们表明,在温和的假设下,FB计划的收敛保证类似于在欧几里得空间的近端梯度算法。

英文摘要:

Wasserstein gradient flows are continuous time dynamics that define curves of steepest descent to minimize an objective function over the space of probability measures (i.e., the Wasserstein space). This objective is typically a divergence w.r.t. a fixed target distribution. In recent years, these continuous time dynamics have been used to study the convergence of machine learning algorithms aiming at approximating a probability distribution. However, the discrete-time behavior of these algorithms might differ from the continuous time dynamics. Besides, although discretized gradient flows have been proposed in the literature, little is known about their minimization power. In this work, we propose a Forward Backward (FB) discretization scheme that can tackle the case where the objective function is the sum of a smooth and a nonsmooth geodesically convex terms. Using techniques from convex optimization and optimal transport, we analyze the FB scheme as a minimization algorithm on the Wasserstein space. More precisely, we show under mild assumptions that the FB scheme has convergence guarantees similar to the proximal gradient algorithm in Euclidean spaces.

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。

↑