Solving ODE with Universal Flows: Approximation Theory for Flow-Based Models

Abstract

Normalizing flows are powerful invertible probabilistic models that can be used to translate two probability distributions, in a way that allows us to efficiently track the change of probability density. However, to trade for computational efficiency in sampling and in evaluating the log-density, special parameterization designs have been proposed at the cost of representational expressiveness. In this work, we propose to use ODEs as a framework to establish universal approximation theory for certain families of flow-based models.

Cite

Text

Huang et al. "Solving ODE with Universal Flows: Approximation Theory for Flow-Based Models." ICLR 2020 Workshops: DeepDiffEq, 2020.

Markdown

[Huang et al. "Solving ODE with Universal Flows: Approximation Theory for Flow-Based Models." ICLR 2020 Workshops: DeepDiffEq, 2020.](https://mlanthology.org/iclrw/2020/huang2020iclrw-solving/)

BibTeX

@inproceedings{huang2020iclrw-solving,
  title     = {{Solving ODE with Universal Flows: Approximation Theory for Flow-Based Models}},
  author    = {Huang, Chin-Wei and Dinh, Laurent and Courville, Aaron},
  booktitle = {ICLR 2020 Workshops: DeepDiffEq},
  year      = {2020},
  url       = {https://mlanthology.org/iclrw/2020/huang2020iclrw-solving/}
}