HINT: Hierarchical Invertible Neural Transport for Density Estimation and Bayesian Inference

Jakob Kruse, Gianluca Detommaso, Robert Scheichl, Ullrich Köthe

Research output: Contribution to journalArticlepeer-review

Abstract

A large proportion of recent invertible neural architectures is based on a coupling block design. It operates by dividing incoming variables into two sub-spaces, one of which parameterizes an easily invertible (usually affine) transformation that is applied to the other. While the Jacobian of such a transformation is triangular, it is very sparse and thus may lack expressiveness. This work presents a simple remedy by noting that (affine) coupling can be repeated recursively within the resulting sub-spaces, leading to an efficiently invertible block with dense triangular Jacobian. By formulating our recursive coupling scheme via a hierarchical architecture, HINT allows sampling from a joint distribution p(y,x) and the corresponding posterior p(x|y) using a single invertible network. We demonstrate the power of our method for density estimation and Bayesian inference on a novel data set of 2D shapes in Fourier parameterization, which enables consistent visualization of samples for different dimensionalities.
Original languageEnglish
Pages (from-to)8191-8199
JournalProceedings of the 35th AAAI Conference on Artificial Intelligence (AAAI-21)
Volume35
Issue number9
DOIs
Publication statusPublished - 18 May 2021

Keywords

  • stat.ML
  • cs.AI
  • cs.LG

Fingerprint

Dive into the research topics of 'HINT: Hierarchical Invertible Neural Transport for Density Estimation and Bayesian Inference'. Together they form a unique fingerprint.

Cite this