Improved supervised learning methods for EoR parameters reconstruction

Abstract

Within the next few years, the Square Kilometer Array (SKA) or one of its pathfinders will hopefully provide a detection of the 21-cm signal fluctuations from the Epoch of Reionization (EoR). Then, the main goal will be to accurately constrain the underlying astrophysical parameters. Currently, this is mainly done with Bayesian inference using Markov Chain Monte Carlo sampling. Recently, studies using neural networks trained to performed inverse modelling have shown interesting results. We build on these by improving the accuracy of the predictions using neural network and exploring other supervised learning methods: the kernel and ridge regressions. Based on a large training set of 21-cm power spectra, we compare the performances of these supervised learning methods. When using an un-noised signal as input, we improve on previous neural network accuracy by one order of magnitude and, using local ridge kernel regression, we gain another factor of a few. We then reach a rms prediction error of a few percents of the 1-sigma confidence level due to SKA thermal noise (as estimated with Bayesian inference). This last performance level requires optimizing the hyper-parameters of the method: how to do that perfectly in the case of an unknown signal remains an open question. For an input signal altered by a SKA-type thermal noise, our neural network recovers the astrophysical parameter values with an error within half of the 1$\sigma$ confidence level due to the SKA thermal noise. This accuracy improves to 10$\%$ of the 1$\sigma$ level when using the local ridge kernel regression (with optimized hyper-parameters). We are thus reaching a performance level where supervised learning methods are a viable alternative to determine the best-fit parameters values.

arXiv comment14 pages, 6 figures, 4 tables, submitted to MNRAS

References in corpus (39)

• Cosmology at Low Frequencies: The 21 cm Transition and the High-Redshift Universe
• An absorption profile centred at 78 megahertz in the sky-averaged spectrum
• 21-cm cosmology
• 21-cm Tomography of the Intergalactic Medium at High Redshift
• 21cmFAST: A Fast, Semi-Numerical Simulation of the High-Redshift 21-cm Signal
• Cosmological Parameter Estimation Using 21 cm Radiation from the Epoch of Reionization
• Efficient Simulations of Early Structure Formation and Reionization
• The morphology of HII regions during reionization
• What Next-Generation 21 cm Power Spectrum Measurements Can Teach Us About the Epoch of Reionization
• Reionization and the Cosmic Dawn with the Square Kilometre Array
• A simulation calibrated limit on the HI power spectrum from the GMRT Epoch of Reionization experiment
• Upper limits on the 21-cm Epoch of Reionization power spectrum from one night with LOFAR
• RadioAstron -- a Telescope with a Size of 300 000 km: Main Parameters and First Observational Results
• 21 cm fluctuations from inhomogeneous X-ray heating before reionization
• Foreground simulations for the LOFAR - Epoch of Reionization Experiment
• Inferring the astrophysics of reionization and cosmic dawn from galaxy luminosity functions and the 21-cm signal
• A Method for Separating the Physics from the Astrophysics of High-Redshift 21cm Fluctuations
• 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
• Statistical Probes of Reionization With 21 cm Tomography
• The observable signature of late heating of the Universe during cosmic reionization
• First Season MWA EoR Power Spectrum Results at Redshift 7
• Fast Large Volume Simulations of the 21 cm Signal from the Reionization and pre-Reionization Epochs
• The Cosmic Dawn and Epoch of Reionization with the Square Kilometre Array
• Redshifted 21 cm Emission From the Pre-Reionization Era I. Mean Signal and Linear Fluctuations
• 21 cm signal from cosmic dawn: Imprints of spin temperature fluctuations and peculiar velocities
• The 21 centimeter emission from the reionization epoch: extended and point source foregrounds
• 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
• Fast Large-Scale Reionization Simulations
• The simulated 21 cm signal during the epoch of reionization : full modeling of the Ly-alpha pumping
• 21CMMC with a 3D light-cone: the impact of the co-evolution approximation on the astrophysics of reionisation and cosmic dawn
• Analysing the 21cm signal from the Epoch of Reionization with artificial neural networks
• Emulation of reionization simulations for Bayesian inference of astrophysics parameters using neural networks
• Evaluating machine learning techniques for predicting power spectra from reionization simulations
• A method to determine the evolution history of the mean neutral Hydrogen fraction
• Simultaneously constraining the astrophysics of reionisation and the epoch of heating with 21CMMC
• Radio views of cosmic reionization
• Working towards an optimal sampling of the 21 cm signal parameter space

Cited by in corpus (17)

• Simulation-Based Inference of Reionization Parameters From 3D Tomographic 21 cm Lightcone Images
• Parameters Estimation for the Cosmic Microwave Background with Bayesian Neural Networks
• Implicit Likelihood Inference of Reionization Parameters from the 21 cm Power Spectrum
• Machine Learning for Observational Cosmology
• The impact of non-Gaussianity on the Epoch of Reionization parameter forecast using 21-cm power spectrum measurements
• A unified framework for 21cm tomography sample generation and parameter inference with Progressively Growing GANs
• Detecting the non-Gaussianity of the 21-cm signal during reionisation with the Wavelet Scattering Transform
• Accurate modelling of the Lyman-$\alpha$ coupling for the 21-cm signal, observability with NenuFAR and SKA
• SKA-Low Intensity Mapping Pathfinder Updates: Deeper 21 cm Power Spectrum Limits from Improved Analysis Frameworks
• A fast method of reionization parameter space exploration using GPR trained SCRIPT
• Efficient exploration of reionization parameters for the upcoming 21 cm observations using a photon conserving semi-numerical model SCRIPT
• Predicting 21cm-line map from Lyman $\alpha$ emitter distribution with Generative Adversarial Networks
• Deconfusing intensity maps with neural networks
• Probing the Inflaton Potential with SKA
• Nonlinear reconstruction of 21cm global signal from 21cm power spectrum with artificial neural networks
• Accelerating reionization constraints: An ANN-emulator framework for the SCRIPT Semi-numerical Model
• A bubble size distribution model for the Epoch of Reionization