ICLR 2021poster5 citations
Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows
Chris Cannella, Mohammadreza Soltani, Vahid Tarokh
Abstract
We introduce Projected Latent Markov Chain Monte Carlo (PL-MCMC), a technique for sampling from the exact conditional distributions learned by normalizing flows. As a conditional sampling method, PL-MCMC enables Monte Carlo Expectation Maximization (MC-EM) training of normalizing flows from incomplete data. Through experimental tests applying normalizing flows to missing data tasks for a variety of data sets, we demonstrate the efficacy of PL-MCMC for conditional sampling from normalizing flows.
Conditional SamplingNormalizing FlowsMarkov Chain Monte CarloMissing Data Inference
BibTeX
@inproceedings{
cannella2021projected,
title={Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows},
author={Chris Cannella and Mohammadreza Soltani and Vahid Tarokh},
booktitle={International Conference on Learning Representations},
year={2021},
url={https://openreview.net/forum?id=MBpHUFrcG2x}
}