Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows
Publication
, Conference
Cannella, C; Soltani, M; Tarokh, V
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.
Duke Scholars
Conference Name
International Conference on Learning Representation (ICLR)
Citation
APA
Chicago
ICMJE
MLA
NLM
Cannella, C., Soltani, M., & Tarokh, V. (n.d.). Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows. Presented at the International Conference on Learning Representation (ICLR).
Cannella, Chris, Mohammadreza Soltani, and Vahid Tarokh. “Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows,” n.d.
Cannella C, Soltani M, Tarokh V. Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows. In.
Cannella, Chris, et al. Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows.
Cannella C, Soltani M, Tarokh V. Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows.
Conference Name
International Conference on Learning Representation (ICLR)