Stochastic Normalizing Flows
By Liam Hodgkinson, Christopher van der Heide, Fred Roosta, Michael W. Mahoney. 2020
We introduce stochastic normalizing flows, an extension of continuous normalizing flows for maximum likelihood estimation and variational inference (VI) using stochastic differential equations (SDEs). Using the theory of rough paths, the underlying Brownian motion is treated as a latent variable and approximated, enabling efficient training of neural SDEs as random neural ordinary differential equations. These SDEs can be used for constructing efficient Markov chains to sample from the underlying distribution of a given dataset. Furthermore, by considering families of targeted SDEs with prescribed stationary distribution, we can apply VI to the optimization of hyperparameters in stochastic MCMC.
Subfields: Generative Models, Probabilistic & Bayesian Methods
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Generative Models
- Frank No'e
- Jonas Köhler
- Hao Wu
- Steven L. Brunton
- Leon Klein
- Qinsheng Zhang
- Yongxin Chen
- Ricardo Vinuesa
In Probabilistic & Bayesian Methods
- Frank No'e
- Jonas Köhler
- Hao Wu
- Steven L. Brunton
- Leon Klein
- Ricardo Vinuesa
- Tommi S. Jaakkola
- Regina Barzilay
Citations in the corpus
- Stochastic Normalizing Flows (2020, background): “Furthermore related is the work of (Sohl-Dickstein et al., 2015; Chen et al., 2017; Gu et al., 2019; Hodgkinson et al., 2020) which all propose learnable stochastic processes designed for density estimation and/or sampling.”