TY - JOUR
T1 - Iterative refinement of the approximate posterior for directed belief networks
AU - Hjelm, R. Devon
AU - Cho, Kyunghyun
AU - Chung, Junyoung
AU - Salakhutdinov, Russ
AU - Calhoun, Vince
AU - Jojic, Nebojsa
N1 - Funding Information:
This work was supported by Microsoft Research to RDH under NJ; NIH P20GM103472, R01 grant REB020407, and NSF grant 1539067 to VDC; and ONR grant N000141512791 and ADeLAIDE grant FA8750-16C-0130-001 to RS. KC was supported in part by Facebook, Google (Google Faculty Award 2016) and NVidia (GPU Center of Excellence 2015-2016), and RDH was supported in part by PIBBS.
Publisher Copyright:
© 2016 NIPS Foundation - All Rights Reserved.
PY - 2016
Y1 - 2016
N2 - Variational methods that rely on a recognition network to approximate the posterior of directed graphical models offer better inference and learning than previous methods. Recent advances that exploit the capacity and flexibility in this approach have expanded what kinds of models can be trained. However, as a proposal for the posterior, the capacity of the recognition network is limited, which can constrain the representational power of the generative model and increase the variance of Monte Carlo estimates. To address these issues, we introduce an iterative refinement procedure for improving the approximate posterior of the recognition network and show that training with the refined posterior is competitive with state-of-the-art methods. The advantages of refinement are further evident in an increased effective sample size, which implies a lower variance of gradient estimates.
AB - Variational methods that rely on a recognition network to approximate the posterior of directed graphical models offer better inference and learning than previous methods. Recent advances that exploit the capacity and flexibility in this approach have expanded what kinds of models can be trained. However, as a proposal for the posterior, the capacity of the recognition network is limited, which can constrain the representational power of the generative model and increase the variance of Monte Carlo estimates. To address these issues, we introduce an iterative refinement procedure for improving the approximate posterior of the recognition network and show that training with the refined posterior is competitive with state-of-the-art methods. The advantages of refinement are further evident in an increased effective sample size, which implies a lower variance of gradient estimates.
UR - http://www.scopus.com/inward/record.url?scp=85018918401&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85018918401&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85018918401
SN - 1049-5258
SP - 4698
EP - 4706
JO - Advances in Neural Information Processing Systems
JF - Advances in Neural Information Processing Systems
T2 - 30th Annual Conference on Neural Information Processing Systems, NIPS 2016
Y2 - 5 December 2016 through 10 December 2016
ER -