Timezone: »

 
Action Matching: A Variational Method for Learning Stochastic Dynamics from Samples
Kirill Neklyudov · Daniel Severo · Alireza Makhzani

@
in NeurIPS 2022 Workshop on Score-Based Methods »
Event URL: https://openreview.net/forum?id=rXiZMBJBdB »

Stochastic dynamics are ubiquitous in many fields of science, from the evolution of quantum systems in physics to diffusion-based models in machine learning. Existing methods such as score matching can be used to simulate these physical processes by assuming that the dynamics is a diffusion, which is not always the case. In this work, we propose a method called "Action Matching" that enables us to learn a much broader family of stochastic dynamics. Our method requires access only to samples from different time-steps, makes no explicit assumptions about the underlying dynamics, and can be applied even when samples are uncorrelated (i.e., are not part of a trajectory). Action Matching directly learns the underlying mechanism that moves samples in time without modeling the distributions at each time-step. In this work, we showcase how Action Matching can be used for generative modeling for computer vision tasks and discuss potential applications in other areas of science.

Author Information

Kirill Neklyudov (Vector Institute)
Daniel Severo (University of Toronto)
Alireza Makhzani (University of Toronto)

More from the Same Authors