Large-Scale Wasserstein Gradient Flows

Petr Mokrov, Alexander Korotin, Lingxiao Li, Aude Genevay, Justin Solomon, Evgeny Burnaev

Keywords: wasserstein-2 gradient flows, JKO stepping, optimal transport, input-convex neural networks, diffusion processes

TL;DR: We propose a scalable method to approximate diffusion processes via Wasserstein-2 gradient flows empowered by input-convex neural networks and Jordan-Kinderlehrer-Otto time discretization.

Abstract: Wasserstein gradient flows provide a powerful means of understanding and solving many diffusion equations. Specifically, Fokker-Planck equations, which model the diffusion of probability measures, can be understood as gradient descent over entropy functionals in Wasserstein space. This equivalence, introduced by Jordan, Kinderlehrer and Otto, inspired the so-called JKO scheme to approximate these diffusion processes via an implicit discretization of the gradient flow in Wasserstein space. Solving the optimization problem associated with each JKO step, however, presents serious computational challenges. We introduce a scalable method to approximate Wasserstein gradient flows, targeted to machine learning applications. Our approach relies on input-convex neural networks (ICNNs) to discretize the JKO steps, which can be optimized by stochastic gradient descent. Contrarily to previous work, our method does not require domain discretization or particle simulation. As a result, we can sample from the measure at each time step of the diffusion and compute its probability density. We demonstrate the performance of our algorithm by computing diffusions following the Fokker-Planck equation and apply it to unnormalized density sampling as well as nonlinear filtering.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

Code: https://github.com/PetrMokrov/Large-Scale-Wasserstein-Gradient-Flows

Community Implementations: [ 2 code implementations](https://www.catalyzex.com/paper/large-scale-wasserstein-gradient-flows/code)