Faithful inversion of generative models for effective amortized inference
Inference amortization methods share information across multiple posteriorinference problems, allowing each to be carried out more efficiently. Generally, they require the inversion of the dependency structure in the generative model, as the modeller must learn a mapping from observations to distrib...
Autors principals: | Webb, S, Golinski, A, Zinkov, R, Narayanaswamy, S, Rainforth, T, Teh, Y, Wood, F |
---|---|
Format: | Conference item |
Publicat: |
Curran Associates
2019
|
Ítems similars
-
Amortized Monte Carlo integration
per: Goliński, A, et al.
Publicat: (2019) -
Amortized rejection sampling in universal probabilistic programming
per: Naderiparizi, S, et al.
Publicat: (2022) -
Amortized inference and model learning for probabilistic programming
per: Le, TA
Publicat: (2019) -
Kilonova Spectral Inverse Modelling with Simulation-based Inference: An Amortized Neural Posterior Estimation Analysis
per: P. Darc, et al.
Publicat: (2024-01-01) -
Target–aware Bayesian inference: how to beat optimal conventional estimators
per: Rainforth, T, et al.
Publicat: (2020)