Learning cosmology and clustering with cosmic graphs

Abstract

We train deep learning models on thousands of galaxy catalogues from the state-of-the-art hydrodynamic simulations of the CAMELS project to perform regression and inference. We employ Graph Neural Networks (GNNs), architectures designed to work with irregular and sparse data, like the distribution of galaxies in the Universe. We first show that GNNs can learn to compute the power spectrum of galaxy catalogues with a few percent accuracy. We then train GNNs to perform likelihood-free inference at the galaxy-field level. Our models are able to infer the value of ${\Omega }_{\mathrm{m}}$ with a $\sim 12\%-13\%$ accuracy just from the positions of $\sim 1000$ galaxies in a volume of $(25h^{-1}\mathrm{Mpc}{)}^3$ at $z=0$ while accounting for astrophysical uncertainties as modelled in CAMELS. Incorporating information from galaxy properties, such as stellar mass, stellar metallicity, and stellar radius, increases the accuracy to $4\%-8\%$. Our models are built to be translational and rotational invariant, and they can extract information from any scale larger than the minimum distance between two galaxies. However, our models are not completely robust: testing on simulations run with a different subgrid physics than the ones used for training does not yield as accurate results.

21 pages, 8 figures, code publicly available at https://github.com/PabloVD/CosmoGraphNet

References in corpus (10)

• Simulating galaxy formation with black hole driven thermal and kinetic feedback
• Simba: Cosmological Simulations with Black Hole Growth and Feedback
• Constraining $M_{\nu }$ with the Bispectrum II: The Total Information Content of the Galaxy Bispectrum
• Constraining neutrino mass with weak lensing Minkowski Functionals
• A Hybrid Deep Learning Approach to Cosmological Constraints From Galaxy Redshift Surveys
• Cosmology with one galaxy?
• Breaking baryon-cosmology degeneracy with the electron density power spectrum
• Minkowski Functionals of Convergence Maps and the Lensing Figure of Merit
• Extracting cosmological parameters from N-body simulations using machine learning techniques
• Probing massive neutrinos with the Minkowski functionals of large-scale structure

Cited by in corpus (5)

• Field Level Neural Network Emulator for Cosmological N-body Simulations
• Large-scale density and velocity field reconstructions with neural networks
• Robust Neural Network-Enhanced Estimation of Local Primordial Non-Gaussianity
• Machine-learning cosmology from void properties
• A graph model for the clustering of dark matter halos