Evaluating machine learning techniques for predicting power spectra from reionization simulations

Abstract

Upcoming experiments such as the SKA will provide huge quantities of data. Fast modelling of the high-redshift 21cm signal will be crucial for efficiently comparing these data sets with theory. The most detailed theoretical predictions currently come from numerical simulations and from faster but less accurate semi-numerical simulations. Recently, machine learning techniques have been proposed to emulate the behaviour of these semi-numerical simulations with drastically reduced time and computing cost. We compare the viability of five such machine learning techniques for emulating the 21cm power spectrum of the publicly-available code SimFast21. Our best emulator is a multilayer perceptron with three hidden layers, reproducing SimFast21 power spectra $10^8$ times faster than the simulation with 4% mean squared error averaged across all redshifts and input parameters. The other techniques (interpolation, Gaussian processes regression, and support vector machine) have slower prediction times and worse prediction accuracy than the multilayer perceptron. All our emulators can make predictions at any redshift and scale, which gives more flexible predictions but results in significantly worse prediction accuracy at lower redshifts. We then present a proof-of-concept technique for mapping between two different simulations, exploiting our best emulator's fast prediction speed. We demonstrate this technique to find a mapping between SimFast21 and another publicly-available code 21cmFAST. We observe a noticeable offset between the simulations for some regions of the input space. Such techniques could potentially be used as a bridge between fast semi-numerical simulations and accurate numerical radiative transfer simulations.

arXiv comment17 pages, 14 figures

References in corpus (40)

• Adam: A Method for Stochastic Optimization
• Scikit-learn: Machine Learning in Python
• Ellipsoidal collapse and an improved model for the number and spatial distribution of dark matter haloes
• Cosmology at Low Frequencies: The 21 cm Transition and the High-Redshift Universe
• In the Beginning: The First Sources of Light and the Reionization of the Universe
• The Murchison Widefield Array: the Square Kilometre Array Precursor at low radio frequencies
• 21-cm cosmology
• 21cmFAST: A Fast, Semi-Numerical Simulation of the High-Redshift 21-cm Signal
• Hydrogen Epoch of Reionization Array (HERA)
• The Growth of HII Regions During Reionization
• A Modified Magnitude System that Produces Well-Behaved Magnitudes, Colors, and Errors Even for Low Signal-to-Noise Ratio Measurements
• Efficient Simulations of Early Structure Formation and Reionization
• The morphology of HII regions during reionization
• Bright Source Subtraction Requirements For Redshifted 21 cm Measurements
• Reionization and the Cosmic Dawn with the Square Kilometre Array
• Simulations and Analytic Calculations of Bubble Growth During Hydrogen Reionization
• Upper limits on the 21-cm Epoch of Reionization power spectrum from one night with LOFAR
• PAPER-64 Constraints on Reionization: The 21cm Power Spectrum at z=8.4
• 21CMMC: an MCMC analysis tool enabling astrophysical parameter studies of the cosmic 21 cm signal
• C^2-Ray: A new method for photon-conserving transport of ionizing radiation
• Inhomogeneous recombinations during cosmic reionization
• Fast Large Volume Simulations of the 21 cm Signal from the Reionization and pre-Reionization Epochs
• Eliminating the optical depth nuisance from the CMB with 21 cm cosmology
• Quantifying the non-Gaussianity in the EoR 21-cm signal through bispectrum
• Emulating Simulations of Cosmic Dawn for 21cm Power Spectrum Constraints on Cosmology, Reionization, and X-ray Heating
• Deep learning from 21-cm tomography of the Cosmic Dawn and Reionization
• Analysing the 21cm signal from the Epoch of Reionization with artificial neural networks
• On the use of semi-numerical simulations in predicting the 21-cm signal from the epoch of reionization
• Emulation of reionization simulations for Bayesian inference of astrophysics parameters using neural networks
• The accuracy of semi-numerical reionization models in comparison with radiative transfer simulations
• The 21cm bispectrum as a probe of non-Gaussianities due to X-ray heating
• 21cm-line bispectrum as method to probe Cosmic Dawn and Epoch of Reionization
• A fast estimator for the bispectrum and beyond - A practical method for measuring non-Gaussianity in 21-cm maps
• Simulating the 21-cm signal from reionisation including non-linear ionisations and inhomogeneous recombinations
• 21SSD: a public database of simulated 21-cm signals from the epoch of reionization
• Epoch of Reionisation 21cm Forecasting From MCMC-Constrained Semi-Numerical Models
• Constraints on the temperature of the intergalactic medium at z=8.4 with 21-cm observations
• Models of the Cosmological 21 cm Signal from the Epoch of Reionization Calibrated with Lyman-alpha and CMB Data
• Simultaneously constraining the astrophysics of reionisation and the epoch of heating with 21CMMC
• Upper Limits on the 21 cm Power Spectrum at z = 5.9 from Quasar Absorption Line Spectroscopy

Cited by in corpus (43)

• Large-scale dark matter simulations
• HERA Phase I Limits on the Cosmic 21-cm Signal: Constraints on Astrophysics and Cosmology During the Epoch of Reionization
• Constraining the intergalactic medium at $z\approx$ 9.1 using LOFAR Epoch of Reionization observations
• Identification of Young Stellar Object candidates in the $Gaia$ DR2 x AllWISE catalogue with machine learning methods
• Tight Constraints on the Excess Radio Background at $z=9.1$ from LOFAR
• Constraining the primordial black hole abundance with 21cm cosmology
• Emulating the Global 21-cm Signal from Cosmic Dawn and Reionization
• Simulation-Based Inference of Reionization Parameters From 3D Tomographic 21 cm Lightcone Images
• Bayesian emulator optimisation for cosmology: application to the Lyman-alpha forest
• Accurate emulator for the redshift-space power spectrum of dark matter halos and its application to galaxy power spectrum
• GLOBALEMU: A novel and robust approach for emulating the sky-averaged 21-cm signal from the cosmic dawn and epoch of reionisation
• Machine learning astrophysics from 21 cm lightcones: impact of network architectures and signal contamination
• General framework for cosmological dark matter bounds using $N$-body simulations
• Implicit Likelihood Inference of Reionization Parameters from the 21 cm Power Spectrum
• Cosmological parameter estimation via iterative emulation of likelihoods
• Deep learning approach for identification of HII regions during reionization in 21-cm observations
• Exploring the likelihood of the 21-cm power spectrum with simulation-based inference
• 21cmEMU: an emulator of 21cmFAST summary observables
• Improved supervised learning methods for EoR parameters reconstruction
• Cosmic Velocity Field Reconstruction Using AI
• Extracting the 21-cm Power Spectrum and the reionization parameters from mock datasets using Artificial Neural Networks
• Using Artificial Neural Networks to extract the 21-cm Global Signal from the EDGES data
• Decoding Dark Matter Substructure without Supervision
• Accurate modelling of the Lyman-$\alpha$ coupling for the 21-cm signal, observability with NenuFAR and SKA
• Extracting the Global 21-cm signal from Cosmic Dawn and Epoch of Reionization in the presence of Foreground and Ionosphere
• The multi-frequency angular power spectrum in parameter studies of the cosmic 21-cm signal
• Parameter inference and model comparison using theoretical predictions from noisy simulations
• Estimation of HII Bubble Size Distribution from 21cm Power Spectrum with Artificial Neural Networks
• Predicting 21cm-line map from Lyman $\alpha$ emitter distribution with Generative Adversarial Networks
• Analysing the Epoch of Reionization with three-point correlation functions and machine learning techniques
• Comparing sampling techniques to chart parameter space of 21 cm Global signal with Artificial Neural Networks
• Comparison of Bayesian inference methods using the Loreli II database of hydro-radiative simulations of the 21-cm signal
• Synthetic Observations with the Square Kilometre Array (SKA) -- development towards an end-to-end pipeline
• Impact of Calibration and Position Errors on Astrophysical Parameters of the HI 21cm Signal
• In Search of Global 21-cm Signal using Artificial Neural Network in light of ARCADE 2
• From ANN to BNN: Inferring Reionization Parameters using Uncertainty-aware Emulators of 21-cm Summaries
• Exploring Machine Learning Regression Models for Advancing Foreground Mitigation and Global 21cm Signal Parameter Extraction
• CosmoUiT: A Vision Transformer-UNet Hybrid for Fast and Accurate Emulation of 21-cm Maps from the Epoch of Reionization
• Accelerating reionization constraints: An ANN-emulator framework for the SCRIPT Semi-numerical Model
• Simulation based inference of the ionization history from the 2D 21 cm power spectrum
• A bubble size distribution model for the Epoch of Reionization
• Nonlinear reconstruction of 21cm global signal from 21cm power spectrum with artificial neural networks
• Neural network emulator to constrain the high-$z$ IGM thermal state from Lyman-$\alpha$ forest flux auto-correlation function