HNPE: Leveraging Global Parameters for Neural Posterior Estimation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

HNPE: Leveraging Global Parameters for Neural Posterior Estimation


Inferring the parameters of a stochastic model based on experimental observations is central to the scientific method. A particularly challenging setting is when the model is strongly indeterminate, i.e. when distinct sets of parameters yield identical observations. This arises in many practical situations, such as when inferring the distance and power of a radio source (is the source close and weak or far and strong?) or when estimating the amplifier gain and underlying brain activity of an electrophysiological experiment. In this work, we present hierarchical neural posterior estimation (HNPE), a novel method for cracking such indeterminacy by exploiting additional information conveyed by an auxiliary set of observations sharing global parameters. Our method extends recent developments in simulation-based inference (SBI) based on normalizing flows to Bayesian hierarchical models. We validate quantitatively our proposal on a motivating example amenable to analytical solutions and then apply it to invert a well known non-linear model from computational neuroscience.
Fichier principal
Vignette du fichier
neurips_2021.pdf (1.83 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03139916 , version 1 (12-02-2021)
hal-03139916 , version 2 (06-10-2021)
hal-03139916 , version 3 (10-11-2021)


  • HAL Id : hal-03139916 , version 3


Pedro Luiz Coelho Rodrigues, Thomas Moreau, Gilles Louppe, Alexandre Gramfort. HNPE: Leveraging Global Parameters for Neural Posterior Estimation. NeurIPS 2021, Dec 2021, Sydney (Online), Australia. ⟨hal-03139916v3⟩
207 View
111 Download


Gmail Facebook X LinkedIn More