Kernel Selection for Gaussian Process in Cosmology: with Approximate Bayesian Computation Rejection and Nested Sampling

Abstract

Gaussian Process (GP) has gained much attention in cosmology due to its ability to reconstruct cosmological data in a model-independent manner. In this study, we compare two methods for GP kernel selection: Approximate Bayesian Computation (ABC) Rejection and nested sampling. We analyze three types of data: cosmic Chronometer data (CC), Type Ia Supernovae (SNIa), and Gamma Ray Burst (GRB), using five kernel functions. To evaluate the differences between kernel functions, we assess the strength of evidence using Bayes factors. Our results show that, for ABC Rejection, the Matérn kernel with $\nu$=5/2 (M52 kernel) outperformes the commonly used Radial Basis Function (RBF) kernel in approximating all three datasets. Bayes factors indicate that the M52 kernel typically supports the observed data better than the RBF kernel, but with no clear advantage over other alternatives. However, nested sampling gives different results, with the M52 kernel losing its advantage. Nevertheless, Bayes factors indicate no significant dependence of the data on each kernel.

15 pages, 4 tables, 6 figures, accepted for publication in ApJS. We draw a new conclusion which is different from the standard ABC-related methods(e.g. ABC-SMC) in kernel selection for Gaussian Process

References in corpus (9)

• Array Programming with NumPy
• Bayes in the sky: Bayesian inference and model selection in cosmology
• Raising the bar: new constraints on the Hubble parameter with cosmic chronometers at z$\sim$2
• PolyChord: nested sampling for cosmology
• Gaussian Process Cosmography
• Cosmology with gamma-ray bursts: I. The Hubble diagram through the calibrated $E_{\mathrm{p},\mathrm{i}}$ - $E_{\mathrm{iso}}$ correlation
• astroABC: An Approximate Bayesian Computation Sequential Monte Carlo sampler for cosmological parameter estimation
• Sequential Monte Carlo with Adaptive Weights for Approximate Bayesian Computation
• Towards a model-independent reconstruction approach for late-time Hubble data