Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Miller, Benjamin Kurt, Weniger, Christoph, Forré, Patrick
Format: Preprint
Veröffentlicht: 2022
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2210.06170
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929409525547008
author Miller, Benjamin Kurt
Weniger, Christoph
Forré, Patrick
author_facet Miller, Benjamin Kurt
Weniger, Christoph
Forré, Patrick
contents Likelihood-to-evidence ratio estimation is usually cast as either a binary (NRE-A) or a multiclass (NRE-B) classification task. In contrast to the binary classification framework, the current formulation of the multiclass version has an intrinsic and unknown bias term, making otherwise informative diagnostics unreliable. We propose a multiclass framework free from the bias inherent to NRE-B at optimum, leaving us in the position to run diagnostics that practitioners depend on. It also recovers NRE-A in one corner case and NRE-B in the limiting case. For fair comparison, we benchmark the behavior of all algorithms in both familiar and novel training regimes: when jointly drawn data is unlimited, when data is fixed but prior draws are unlimited, and in the commonplace fixed data and parameters setting. Our investigations reveal that the highest performing models are distant from the competitors (NRE-A, NRE-B) in hyperparameter space. We make a recommendation for hyperparameters distinct from the previous models. We suggest two bounds on the mutual information as performance metrics for simulation-based inference methods, without the need for posterior samples, and provide experimental results. This version corrects a minor implementation error in $γ$, improving results.
format Preprint
id arxiv_https___arxiv_org_abs_2210_06170
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Contrastive Neural Ratio Estimation for Simulation-based Inference
Miller, Benjamin Kurt
Weniger, Christoph
Forré, Patrick
Machine Learning
Instrumentation and Methods for Astrophysics
High Energy Physics - Phenomenology
Likelihood-to-evidence ratio estimation is usually cast as either a binary (NRE-A) or a multiclass (NRE-B) classification task. In contrast to the binary classification framework, the current formulation of the multiclass version has an intrinsic and unknown bias term, making otherwise informative diagnostics unreliable. We propose a multiclass framework free from the bias inherent to NRE-B at optimum, leaving us in the position to run diagnostics that practitioners depend on. It also recovers NRE-A in one corner case and NRE-B in the limiting case. For fair comparison, we benchmark the behavior of all algorithms in both familiar and novel training regimes: when jointly drawn data is unlimited, when data is fixed but prior draws are unlimited, and in the commonplace fixed data and parameters setting. Our investigations reveal that the highest performing models are distant from the competitors (NRE-A, NRE-B) in hyperparameter space. We make a recommendation for hyperparameters distinct from the previous models. We suggest two bounds on the mutual information as performance metrics for simulation-based inference methods, without the need for posterior samples, and provide experimental results. This version corrects a minor implementation error in $γ$, improving results.
title Contrastive Neural Ratio Estimation for Simulation-based Inference
topic Machine Learning
Instrumentation and Methods for Astrophysics
High Energy Physics - Phenomenology
url https://arxiv.org/abs/2210.06170