Model-independent Gamma-Ray Bursts Constraints on Cosmological Models Using Machine Learning

TL;DR

This study employs machine learning techniques to calibrate gamma-ray burst luminosity relations independently of cosmological models, enabling the use of GRBs for constraining cosmological parameters.

Contribution

It introduces a novel, model-independent calibration method for GRBs using KNN and Random Forest algorithms based on supernova data.

Findings

ML-calibrated GRBs produce consistent cosmological constraints

KNN and RF outperform traditional calibration methods

Results align with Gaussian Process calibration

Abstract

In this paper, we calibrate the luminosity relation of gamma-ray bursts (GRBs) with the machine learning (ML) algorithms from the Pantheon+ sample of type Ia supernovae in a cosmology-independent way. By using K-Nearest Neighbors (KNN) and Random Forest (RF) selected with the best performance in the ML algorithms, we calibrate the Amati relation (\unboldmath{$E_{\rm p}$-$E_{\rm iso}$}) relation with the A219 sample to construct the Hubble diagram of GRBs. Via the Markov Chain Monte Carlo numerical method with GRBs at high redshift and latest observational Hubble data, we find the results of constraints on cosmological models by using KNN and RF algorithms are consistent with those obtained from GRBs calibrated by using the Gaussian Process.