Weak lensing maps contain information beyond two-point statistics on small
scales. Much recent work has tried to extract this information through a range
of different observables or via nonlinear transformations of the lensing field.
Here we train and apply a 2D convolutional neural network to simulated
noiseless lensing maps covering 96 different cosmological models over a range
of $Ømega_m,\sigma_8$. Using the area of the confidence contour in the
$Ømega_m,\sigma_8$ plane as a figure-of-merit, derived from simulated
convergence maps smoothed on a scale of 1.0 arcmin, we show that the neural
network yields $5 \times$ tighter constraints than the power spectrum,
and $4 \times$ tighter than the lensing peaks. Such gains illustrate
the extent to which weak lensing data encode cosmological information not
accessible to the power spectrum or even other, non-Gaussian statistics such as
lensing peaks.
Description
Non-Gaussian information from weak lensing data via deep learning
%0 Generic
%1 gupta2018nongaussian
%A Gupta, Arushi
%A Matilla, José Manuel Zorrilla
%A Hsu, Daniel
%A Haiman, Zoltán
%D 2018
%K cosmology machine_learning
%R 10.1103/PhysRevD.97.103515
%T Non-Gaussian information from weak lensing data via deep learning
%U http://arxiv.org/abs/1802.01212
%X Weak lensing maps contain information beyond two-point statistics on small
scales. Much recent work has tried to extract this information through a range
of different observables or via nonlinear transformations of the lensing field.
Here we train and apply a 2D convolutional neural network to simulated
noiseless lensing maps covering 96 different cosmological models over a range
of $Ømega_m,\sigma_8$. Using the area of the confidence contour in the
$Ømega_m,\sigma_8$ plane as a figure-of-merit, derived from simulated
convergence maps smoothed on a scale of 1.0 arcmin, we show that the neural
network yields $5 \times$ tighter constraints than the power spectrum,
and $4 \times$ tighter than the lensing peaks. Such gains illustrate
the extent to which weak lensing data encode cosmological information not
accessible to the power spectrum or even other, non-Gaussian statistics such as
lensing peaks.
@misc{gupta2018nongaussian,
abstract = {Weak lensing maps contain information beyond two-point statistics on small
scales. Much recent work has tried to extract this information through a range
of different observables or via nonlinear transformations of the lensing field.
Here we train and apply a 2D convolutional neural network to simulated
noiseless lensing maps covering 96 different cosmological models over a range
of {$\Omega_m,\sigma_8$}. Using the area of the confidence contour in the
{$\Omega_m,\sigma_8$} plane as a figure-of-merit, derived from simulated
convergence maps smoothed on a scale of 1.0 arcmin, we show that the neural
network yields $\approx 5 \times$ tighter constraints than the power spectrum,
and $\approx 4 \times$ tighter than the lensing peaks. Such gains illustrate
the extent to which weak lensing data encode cosmological information not
accessible to the power spectrum or even other, non-Gaussian statistics such as
lensing peaks.},
added-at = {2023-05-01T10:39:30.000+0200},
author = {Gupta, Arushi and Matilla, José Manuel Zorrilla and Hsu, Daniel and Haiman, Zoltán},
biburl = {https://www.bibsonomy.org/bibtex/2cbea9545fd257f7cece275746f7dcc15/intfxdx},
description = {Non-Gaussian information from weak lensing data via deep learning},
doi = {10.1103/PhysRevD.97.103515},
interhash = {5b66cddad5c6f9bb83d5f340cf7f7106},
intrahash = {cbea9545fd257f7cece275746f7dcc15},
keywords = {cosmology machine_learning},
note = {cite arxiv:1802.01212Comment: 15 pages, 13 figures, accepted to PRD},
timestamp = {2023-05-01T10:39:30.000+0200},
title = {Non-Gaussian information from weak lensing data via deep learning},
url = {http://arxiv.org/abs/1802.01212},
year = 2018
}