Painting halos from cosmic density fields of dark matter with physically motivated neural networks

Abstract

We present a novel halo painting network that learns to map approximate 3D dark matter fields to realistic halo distributions. This map is provided via a physically motivated network with which we can learn the non-trivial local relation between dark matter density field and halo distributions without relying on a physical model. Unlike other generative or regressive models, a well motivated prior and simple physical principles allow us to train the mapping network quickly and with relatively little data. In learning to paint halo distributions from computationally cheap, analytical and non-linear density fields, we bypass the need for full particle mesh simulations and halo finding algorithms. Furthermore, by design, our halo painting network needs only local patches of dark matter density to predict the halos, and as such, it can predict the 3D halo distribution for any arbitrary simulation box size. Our neural network can be trained using small simulations and used to predict large halo distributions, as long as the resolutions are equivalent. We evaluate our model's ability to generate 3D halo count distributions which reproduce, to a high degree, summary statistics such as the power spectrum and bispectrum, of the input or reference realizations.

arXiv comment18 pages, 14 figures. Accepted for publication in PRD

References in corpus (35)

• Planck 2015 results. XIII. Cosmological parameters
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• The cosmological simulation code GADGET-2
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• The Rockstar Phase-Space Temporal Halo Finder and the Velocity Offsets of Cluster Cores
• Power Spectra for Cold Dark Matter and its Variants
• Multi-scale initial conditions for cosmological simulations
• Unrolled Generative Adversarial Networks
• Bayesian physical reconstruction of initial conditions from large scale structure surveys
• Generative Adversarial Networks recover features in astrophysical images of galaxies beyond the deconvolution limit
• PINOCCHIO: pinpointing orbit-crossing collapsed hierarchical objects in a linear density field
• Learning to Predict the Cosmological Structure Formation
• Physical Bayesian modelling of the non-linear matter distribution: new insights into the Nearby Universe
• Past and present cosmic structure in the SDSS DR7 main sample
• CosmoGAN: creating high-fidelity weak lensing convergence maps using Generative Adversarial Networks
• Predicting the number, spatial distribution and merging history of dark matter haloes
• Fast cosmic web simulations with generative adversarial networks
• Unmasking the Masked Universe: the 2M++ catalogue through Bayesian eyes
• ExoGAN: Retrieving Exoplanetary Atmospheres Using Deep Convolutional Generative Adversarial Networks
• Cosmological Reconstruction From Galaxy Light: Neural Network Based Light-Matter Connection
• Painting with baryons: augmenting N-body simulations with gas using deep generative models
• Machine learning cosmological structure formation
• Classifying the Large Scale Structure of the Universe with Deep Neural Networks
• Cosmological inference from Bayesian forward modelling of deep galaxy redshift surveys
• Deblending galaxy superpositions with branched generative adversarial networks
• A volumetric deep Convolutional Neural Network for simulation of mock dark matter halo catalogues
• Forging new worlds: high-resolution synthetic galaxies with chained generative adversarial networks
• From Dark Matter to Galaxies with Convolutional Networks
• Denoising Weak Lensing Mass Maps with Deep Learning
• LyMAS: Predicting Large-Scale Lyman-alpha Forest Statistics from the Dark Matter Density Field
• HIGAN: Cosmic Neutral Hydrogen with Generative Adversarial Networks
• Generative deep fields: arbitrarily sized, random synthetic astronomical images through deep learning
• PSFGAN: a generative adversarial network system for separating quasar point sources and host galaxy light
• MoLUSC: A MOck Local Universe Survey Constructor

Cited by in corpus (29)

• The Quijote simulations
• Cosmic flows in the nearby Universe: new peculiar velocities from SNe and cosmological constraints
• AI-assisted super-resolution cosmological simulations
• Painting with baryons: augmenting N-body simulations with gas using deep generative models
• Super-resolution emulator of cosmological simulations using deep physical models
• The CAMELS project: public data release
• Learning effective physical laws for generating cosmological hydrodynamics with Lagrangian Deep Learning
• AI-assisted super-resolution cosmological simulations II: Halo substructures, velocities and higher order statistics
• Peculiar velocities in the local Universe: comparison of different models and the implications for $H_0$ and dark matter
• Consistency tests of field level inference with the EFT likelihood
• HIGAN: Cosmic Neutral Hydrogen with Generative Adversarial Networks
• Predicting dark matter halo formation in N-body simulations with deep regression networks
• Simulation-based inference of dynamical galaxy cluster masses with 3D convolutional neural networks
• Nonlinear 3D Cosmic Web Simulation with Heavy-Tailed Generative Adversarial Networks
• Neural physical engines for inferring the halo mass distribution function
• Cosmic Velocity Field Reconstruction Using AI
• A black box for dark sector physics: Predicting dark matter annihilation feedback with conditional GANs
• (DarkAI) Mapping the large-scale density field of dark matter using artificial intelligence
• The cosmic web connection to the dark matter halo distribution through gravity
• Reconstructing dark matter distribution with peculiar velocities: Bayesian forward modelling with corrections for inhomogeneous Malmquist bias
• Halo assembly bias from a deep learning model of halo formation
• Deep learning insights into cosmological structure formation
• Learning neutrino effects in Cosmology with Convolutional Neural Networks
• Weak Lensing Trispectrum and Kurt-Spectra
• A Machine Learning Approach to Correct for Mass Resolution Effects in Simulated Halo Clustering Statistics
• Learning the Universe: Learning to Optimize Cosmic Initial Conditions with Non-Differentiable Structure Formation Models
• Stochastic Super-resolution of Cosmological Simulations with Denoising Diffusion Models
• Deconfusing intensity maps with neural networks
• A New Estimator for Phase Statistics