Evaluating machine learning techniques for predicting power spectra from reionization simulations

@article{Jennings2018EvaluatingML,
  title={Evaluating machine learning techniques for predicting power spectra from reionization simulations},
  author={W D Jennings and Catherine A Watkinson and Filipe B. Abdalla and Jason D. McEwen},
  journal={Monthly Notices of the Royal Astronomical Society},
  year={2018}
}
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… 

Analysing the Epoch of Reionization with three-point correlation functions and machine learning techniques

TLDR
Three-point and high-order clustering statistics of the high-redshift 21cm signal contain valuable information about the Epoch of Reionization that would be particularly useful in regimes where high signal-to-noise ratio prevents direct measurement of bubble size statistics.

Improved supervised learning methods for EoR parameters reconstruction

TLDR
This work improves the accuracy of the predictions using several supervised learning methods: neural networks, kernel regressions, or ridge regressions based on a large training set of 21-cm power spectra, and compares the performances of these methods.

The emulation game : modelling and machine learning for the Epoch of Reionization

TLDR
This thesis presents four projects for efficient modelling and analysis of the results of the Epoch of Reionization (EoR), and fits an analytic clustering model to simulated 21cm maps.

An emulator for the Lyman-α forest

TLDR
It is shown that the Gaussian process emulator has lower emulation error than quadratic polynomial interpolation and thus produces tighter posterior confidence intervals, which will be essential for future Lyman-α surveys such as DESI.

Implicit Likelihood Inference of Reionization Parameters from the 21 cm Power Spectrum

The first measurements of the 21 cm brightness temperature power spectrum from the epoch of reionization will very likely be achieved in the near future by radio interferometric array experiments

Bayesian emulator optimisation for cosmology: application to the Lyman-alpha forest

TLDR
This is the first demonstration of Bayesian optimisation applied to large-scale structure emulation, and it is anticipated the technique will generalise to many other probes such as galaxy clustering, weak lensing and 21cm.

Using artificial neural networks to extract the 21-cm global signal from the EDGES data

The redshifted 21-cm signal of neutral Hydrogen is a promising probe into the period of evolution of our Universe when the first stars were formed (Cosmic Dawn), to the period where the entire

Accurate emulator for the redshift-space power spectrum of dark matter halos and its application to galaxy power spectrum

An accurate theoretical template of the redshift-space galaxy power spectrum, if applicable out to nonlinear scales, enables us to extract more stringent and robust constraints on cosmological

Emulating the global 21-cm signal from Cosmic Dawn and Reionization

The 21-cm signal of neutral hydrogen is a sensitive probe of the Epoch of Reionization (EoR), Cosmic Dawn, and the Dark Ages. Currently, operating radio telescopes have ushered in a data-driven era

Simulation-based Inference of Reionization Parameters from 3D Tomographic 21 cm Light-cone Images

TLDR
This work adopts a trained 3D convolutional neural network to compress the 3D image data into informative summaries (DELFI-3D CNN) and shows that this method recovers accurate posterior distributions for the reionization parameters.

References

SHOWING 1-10 OF 64 REFERENCES

Analysing the 21 cm signal from the epoch of reionization with artificial neural networks

The 21 cm signal from the Epoch of Reionization should be observed within the next decade. While a simple statistical detection is expected with SKA pathfinders, the SKA will hopefully produce a full

21cmfast: a fast, seminumerical simulation of the high‐redshift 21‐cm signal

We introduce a powerful semi-numeric modeling tool, 21cmFAST, designed to efficiently simulate the cosmological 21-cm signal. Our code generates 3D realizations of evolved density, ionization,

On the use of seminumerical simulations in predicting the 21-cm signal from the epoch of reionization

We present a detailed comparison of three different simulations of the epoch of reionization (EoR). The radiative transfer simulation (C-2-RAY) among them is our benchmark. Radiative transfer codes

Epoch of reionization 21 cm forecasting from MCMC-constrained semi-numerical models

The recent low value of Planck (2016) integrated optical depth to Thomson scattering suggests that the reionization occurred fairly suddenly, disfavoring extended reionization scenarios. This will

Emulating Simulations of Cosmic Dawn for 21 cm Power Spectrum Constraints on Cosmology, Reionization, and X-Ray Heating

TLDR
This work uses machine learning algorithms to build a fast emulator that can accurately mimic an expensive simulation of the 21 cm signal across a wide parameter space, and embeds it within a Markov Chain Monte Carlo framework in order to perform Bayesian parameter constraints over a large number of model parameters.

Efficient Simulations of Early Structure Formation and Reionization

Detailed theoretical studies of the high-redshift universe, and especially reionization, are generally forced to rely on time-consuming N-body codes and/or radiative transfer algorithms. We present a

21SSD: a public data base of simulated 21-cm signals from the epoch of reionization

The 21-cm signal from the Epoch of Reionization (EoR) is expected to be detected in the next few years, either with existing instruments or by the upcoming SKA and HERA projects. In this context

Simultaneously constraining the astrophysics of reionisation and the epoch of heating with 21CMMC

Abstract We extend our MCMC sampler of 3D EoR simulations, 21CMMC, to perform parameter estimation directly on light-cones of the cosmic 21cm signal. This brings theoretical analysis one step closer

Deep learning from 21-cm images of the Cosmic Dawn

The 21-cm power spectrum (PS) has been shown to be a powerful discriminant of reionization and cosmic dawn astrophysical parameters. However, the 21-cm tomographic signal is highly non-Gaussian.

The accuracy of semi-numerical reionization models in comparison with radiative transfer simulations

We have developed a modular semi-numerical code that computes the time and spatially dependent ionization of neutral hydrogen (HI), neutral (HeI) and singly ionized helium (HeII) in the intergalactic
...