An improved cosmological parameter inference scheme motivated by deep learning

Dezső Ribli, Bálint Ármin Pataki, I. Csabai

Research output: Contribution to journalLetter

3 Citations (Scopus)

Abstract

Dark matter cannot be observed directly, but its weak gravitational lensing slightly distorts the apparent shapes of background galaxies, making weak lensing one of the most promising probes of cosmology. Several observational studies have measured the effect, and there are currently running1,2 and planned efforts3,4 to provide even larger and higher-resolution weak lensing maps. Owing to nonlinearities on small scales, the traditional analysis with two-point statistics does not fully capture all of the underlying information5. Multiple inference methods have been proposed to extract more details based on higher-order statistics6,7, peak statistics8–13, Minkowski functionals14–16 and recently convolutional neural networks17,18. Here we present an improved convolutional neural network that gives significantly better estimates of the Ωm and σ8 cosmological parameters from simulated weak lensing convergence maps than state-of-art methods and that is also free of systematic bias. We show that the network exploits information in the gradients around peaks, and with this insight we have constructed an easy-to-understand and robust peak-counting algorithm based on the steepness of peaks, instead of their heights. The proposed scheme is even more accurate than the neural network on high-resolution noiseless maps. With shape noise and lower resolution, its relative advantage deteriorates, but it remains more accurate than peak counting.

Original languageEnglish
Pages (from-to)93-98
Number of pages6
JournalNature Astronomy
Volume3
Issue number1
DOIs
Publication statusPublished - Jan 1 2019

Fingerprint

inference
learning
counting
high resolution
cosmology
dark matter
nonlinearity
statistics
slopes
galaxies
gradients
probes
estimates

ASJC Scopus subject areas

  • Astronomy and Astrophysics

Cite this

An improved cosmological parameter inference scheme motivated by deep learning. / Ribli, Dezső; Pataki, Bálint Ármin; Csabai, I.

In: Nature Astronomy, Vol. 3, No. 1, 01.01.2019, p. 93-98.

Research output: Contribution to journalLetter

Ribli, Dezső ; Pataki, Bálint Ármin ; Csabai, I. / An improved cosmological parameter inference scheme motivated by deep learning. In: Nature Astronomy. 2019 ; Vol. 3, No. 1. pp. 93-98.
@article{0760a6dd89af47cf9cf44b58249f1025,
title = "An improved cosmological parameter inference scheme motivated by deep learning",
abstract = "Dark matter cannot be observed directly, but its weak gravitational lensing slightly distorts the apparent shapes of background galaxies, making weak lensing one of the most promising probes of cosmology. Several observational studies have measured the effect, and there are currently running1,2 and planned efforts3,4 to provide even larger and higher-resolution weak lensing maps. Owing to nonlinearities on small scales, the traditional analysis with two-point statistics does not fully capture all of the underlying information5. Multiple inference methods have been proposed to extract more details based on higher-order statistics6,7, peak statistics8–13, Minkowski functionals14–16 and recently convolutional neural networks17,18. Here we present an improved convolutional neural network that gives significantly better estimates of the Ωm and σ8 cosmological parameters from simulated weak lensing convergence maps than state-of-art methods and that is also free of systematic bias. We show that the network exploits information in the gradients around peaks, and with this insight we have constructed an easy-to-understand and robust peak-counting algorithm based on the steepness of peaks, instead of their heights. The proposed scheme is even more accurate than the neural network on high-resolution noiseless maps. With shape noise and lower resolution, its relative advantage deteriorates, but it remains more accurate than peak counting.",
author = "Dezső Ribli and Pataki, {B{\'a}lint {\'A}rmin} and I. Csabai",
year = "2019",
month = "1",
day = "1",
doi = "10.1038/s41550-018-0596-8",
language = "English",
volume = "3",
pages = "93--98",
journal = "Nature Astronomy",
issn = "2397-3366",
publisher = "Nature Publishing Group",
number = "1",

}

TY - JOUR

T1 - An improved cosmological parameter inference scheme motivated by deep learning

AU - Ribli, Dezső

AU - Pataki, Bálint Ármin

AU - Csabai, I.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Dark matter cannot be observed directly, but its weak gravitational lensing slightly distorts the apparent shapes of background galaxies, making weak lensing one of the most promising probes of cosmology. Several observational studies have measured the effect, and there are currently running1,2 and planned efforts3,4 to provide even larger and higher-resolution weak lensing maps. Owing to nonlinearities on small scales, the traditional analysis with two-point statistics does not fully capture all of the underlying information5. Multiple inference methods have been proposed to extract more details based on higher-order statistics6,7, peak statistics8–13, Minkowski functionals14–16 and recently convolutional neural networks17,18. Here we present an improved convolutional neural network that gives significantly better estimates of the Ωm and σ8 cosmological parameters from simulated weak lensing convergence maps than state-of-art methods and that is also free of systematic bias. We show that the network exploits information in the gradients around peaks, and with this insight we have constructed an easy-to-understand and robust peak-counting algorithm based on the steepness of peaks, instead of their heights. The proposed scheme is even more accurate than the neural network on high-resolution noiseless maps. With shape noise and lower resolution, its relative advantage deteriorates, but it remains more accurate than peak counting.

AB - Dark matter cannot be observed directly, but its weak gravitational lensing slightly distorts the apparent shapes of background galaxies, making weak lensing one of the most promising probes of cosmology. Several observational studies have measured the effect, and there are currently running1,2 and planned efforts3,4 to provide even larger and higher-resolution weak lensing maps. Owing to nonlinearities on small scales, the traditional analysis with two-point statistics does not fully capture all of the underlying information5. Multiple inference methods have been proposed to extract more details based on higher-order statistics6,7, peak statistics8–13, Minkowski functionals14–16 and recently convolutional neural networks17,18. Here we present an improved convolutional neural network that gives significantly better estimates of the Ωm and σ8 cosmological parameters from simulated weak lensing convergence maps than state-of-art methods and that is also free of systematic bias. We show that the network exploits information in the gradients around peaks, and with this insight we have constructed an easy-to-understand and robust peak-counting algorithm based on the steepness of peaks, instead of their heights. The proposed scheme is even more accurate than the neural network on high-resolution noiseless maps. With shape noise and lower resolution, its relative advantage deteriorates, but it remains more accurate than peak counting.

UR - http://www.scopus.com/inward/record.url?scp=85059758904&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059758904&partnerID=8YFLogxK

U2 - 10.1038/s41550-018-0596-8

DO - 10.1038/s41550-018-0596-8

M3 - Letter

VL - 3

SP - 93

EP - 98

JO - Nature Astronomy

JF - Nature Astronomy

SN - 2397-3366

IS - 1

ER -