Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userLikelihood-free Cosmological Constraints with Artificial Neural Networks: An Application on Hubble Parameters and SNe Ia
107
0
0.0
(
0
)
Authors
Yu-Chen Wang
Ask ChatGPT about the research
No Arabic abstract
The errors of cosmological data generated from complex processes, such as the observational Hubble parameter data (OHD) and the Type Ia supernova (SN Ia) data, cannot be accurately modeled by simple analytical probability distributions, e.g. Gaussian distribution. To constrain cosmological parameters from these data, likelihood-free inference is usually used to bypass the direct calculation of the likelihood. In this paper, we propose a new procedure to perform likelihood-free cosmological inference using two artificial neural networks (ANN), the Masked Autoregressive Flow (MAF) and the denoising autoencoder (DAE). Our procedure is the first to use DAE to extract features from data, in order to simplify the structure of MAF needed to estimate the posterior. Tested on simulated Hubble parameter data with a simple Gaussian likelihood, the procedure shows the capability of extracting features from data and estimating posterior distributions without the need of tractable likelihood. We demonstrate that it can accurately approximate the real posterior, achieve performance comparable to the traditional MCMC method, and the MAF gets better training results for small number of simulation when the DAE is added. We also discuss the application of the proposed procedure to OHD and Pantheon SN Ia data, and use them to constrain cosmological parameters from the non-flat $Lambda$CDM model. For SNe Ia, we use fitted light curve parameters to find constraints on $H_0,Omega_m,Omega_Lambda$ similar to relevant work, using less empirical distributions. In addition, this work is also the first to use Gaussian process in the procedure of OHD simulation.
rate research
Read More
In this work, we present a new method to estimate cosmological parameters accurately based on the artificial neural network (ANN), and a code called ECoPANN (Estimating Cosmological Parameters with ANN) is developed to achieve parameter inference. We test the ANN method by estimating the basic parameters of the concordance cosmological model using the simulated temperature power spectrum of the cosmic microwave background (CMB). The results show that the ANN performs excellently on best-fit values and errors of parameters, as well as correlations between parameters when compared with that of the Markov Chain Monte Carlo (MCMC) method. Besides, for a well-trained ANN model, it is capable of estimating parameters for multiple experiments that have different precisions, which can greatly reduce the consumption of time and computing resources for parameter inference. Furthermore, we extend the ANN to a multibranch network to achieve a joint constraint on parameters. We test the multibranch network using the simulated temperature and polarization power spectra of the CMB, Type Ia supernovae, and baryon acoustic oscillations, and almost obtain the same results as the MCMC method. Therefore, we propose that the ANN can provide an alternative way to accurately and quickly estimate cosmological parameters, and ECoPANN can be applied to the research of cosmology and even other broader scientific fields.
We use the newly published 28 observational Hubble parameter data ($H(z)$) and current largest SNe Ia samples (Union2.1) to test whether the universe is transparent. Three cosmological-model-independent methods (nearby SNe Ia method, interpolation method and smoothing method) are proposed through comparing opacity-free distance modulus from Hubble parameter data and opacity-dependent distance modulus from SNe Ia . Two parameterizations, $tau(z)=2epsilon z$ and $tau(z)=(1+z)^{2epsilon}-1$ are adopted for the optical depth associated to the cosmic absorption. We find that the results are not sensitive to the methods and parameterizations. Our results support a transparent universe.
The Hubble constant ($H_0$) tension between Type Ia Supernovae (SNe Ia) and Planck measurements ranges from 4 to 6 $sigma$. To investigate this tension, we estimate $H_{0}$ in the $Lambda$CDM and $w_{0}w_{a}$CDM models by dividing the Pantheon sample, the largest compilation of SNe Ia, into 3, 4, 20 and 40 bins. We fit the extracted $H_{0}$ values with a function mimicking the redshift evolution: $g(z)={H_0}(z)=tilde{H}_0/(1+z)^alpha$, where $alpha$ indicates an evolutionary parameter and $tilde{H}_0=H_0$ at $z=0$. We set the absolute magnitude of SNe Ia so that $H_0=73.5,, textrm{km s}^{-1},textrm{Mpc}^{-1}$, and we fix fiducial values for $Omega_{0m}^{Lambda CDM}=0.298$ and $Omega_{0m}^{w_{0}w_{a}CDM}=0.308$. We find that $H_0$ evolves with redshift, showing a slowly decreasing trend, with $alpha$ coefficients consistent with zero only from 1.2 to 2.0 $sigma$. Although the $alpha$ coefficients are compatible with 0 in 3 $sigma$, this however may affect cosmological results. We measure locally a variation of $H_0(z=0)-H_0(z=1)=0.4, textrm{km s}^{-1},textrm{Mpc}^{-1}$ in 3 and 4 bins. Extrapolating ${H_0}(z)$ to $z=1100$, the redshift of the last scattering surface, we obtain values of $H_0$ compatible in 1 $sigma$ with Planck measurements independently of cosmological models and number of bins we investigated. Thus, we have reduced the $H_0$ tension from $54%$ to $72%$ for the $Lambda$CDM and $w_{0}w_{a}$CDM models, respectively. If the decreasing trend of $H_0(z)$ is real, it could be due to astrophysical selection effects or to modified gravity.
We use the Risaliti & Lusso (2015) compilation of 808 X-ray and UV flux measurements of quasars (QSOs) in the redshift range $0.061 leq z leq 6.28$, alone and in conjuction with baryon acoustic oscillation (BAO) and Hubble parameter [$H(z)$] measurements, to constrain cosmological parameters in six cosmological models. The QSO data constraints are significantly weaker than, but consistent with, those from the $H(z)$ + BAO data. A joint analysis of the QSO + $H(z)$ + BAO data is consistent with the current standard model, spatially-flat $Lambda$CDM, but mildly favors closed spatial hypersurfaces and dynamical dark energy.
We present the first cosmological parameter constraints using measurements of type Ia supernovae (SNe Ia) from the Dark Energy Survey Supernova Program (DES-SN). The analysis uses a subsample of 207 spectroscopically confirmed SNe Ia from the first three years of DES-SN, combined with a low-redshift sample of 122 SNe from the literature. Our DES-SN3YR result from these 329 SNe Ia is based on a series of companion analyses and improvements covering SN Ia discovery, spectroscopic selection, photometry, calibration, distance bias corrections, and evaluation of systematic uncertainties. For a flat LCDM model we find a matter density Omega_m = 0.331 +_ 0.038. For a flat wCDM model, and combining our SN Ia constraints with those from the cosmic microwave background (CMB), we find a dark energy equation of state w = -0.978 +_ 0.059, and Omega_m = 0.321 +_ 0.018. For a flat w0waCDM model, and combining probes from SN Ia, CMB and baryon acoustic oscillations, we find w0 = -0.885 +_ 0.114 and wa = -0.387 +_ 0.430. These results are in agreement with a cosmological constant and with previous constraints using SNe Ia (Pantheon, JLA).
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
