Conditional diffusions for amortized neural posterior estimation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Tianyu, Bansal, Vansh, Scott, James G.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909535985205248
author Chen, Tianyu
Bansal, Vansh
Scott, James G.
author_facet Chen, Tianyu
Bansal, Vansh
Scott, James G.
contents Neural posterior estimation (NPE), a simulation-based computational approach for Bayesian inference, has shown great success in approximating complex posterior distributions. Existing NPE methods typically rely on normalizing flows, which approximate a distribution by composing many simple, invertible transformations. But flow-based models, while state of the art for NPE, are known to suffer from several limitations, including training instability and sharp trade-offs between representational power and computational cost. In this work, we demonstrate the effectiveness of conditional diffusions coupled with high-capacity summary networks for amortized NPE. Conditional diffusions address many of the challenges faced by flow-based methods. Our results show that, across a highly varied suite of benchmarking problems for NPE architectures, diffusions offer improved stability, superior accuracy, and faster training times, even with simpler, shallower models. Building on prior work on diffusions for NPE, we show that these gains persist across a variety of different summary network architectures. Code is available at https://github.com/TianyuCodings/cDiff.
format Preprint
id arxiv_https___arxiv_org_abs_2410_19105
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Conditional diffusions for amortized neural posterior estimation
Chen, Tianyu
Bansal, Vansh
Scott, James G.
Machine Learning
Artificial Intelligence
Applications
Neural posterior estimation (NPE), a simulation-based computational approach for Bayesian inference, has shown great success in approximating complex posterior distributions. Existing NPE methods typically rely on normalizing flows, which approximate a distribution by composing many simple, invertible transformations. But flow-based models, while state of the art for NPE, are known to suffer from several limitations, including training instability and sharp trade-offs between representational power and computational cost. In this work, we demonstrate the effectiveness of conditional diffusions coupled with high-capacity summary networks for amortized NPE. Conditional diffusions address many of the challenges faced by flow-based methods. Our results show that, across a highly varied suite of benchmarking problems for NPE architectures, diffusions offer improved stability, superior accuracy, and faster training times, even with simpler, shallower models. Building on prior work on diffusions for NPE, we show that these gains persist across a variety of different summary network architectures. Code is available at https://github.com/TianyuCodings/cDiff.
title Conditional diffusions for amortized neural posterior estimation
topic Machine Learning
Artificial Intelligence
Applications
url https://arxiv.org/abs/2410.19105