Inferring the redshift of more than 150 GRBs with a Machine Learning Ensemble model

• DOI: 10.48550/arxiv.2401.03589
• Publication Date: 2024-01-01
• Authors: Maria Giovanna Da...; Elias Taira; Eric Wang; Elias Lehman; Aditya Narendra; A. Pollo; G. Madejski; V. Petrosian; Małgorzata Bogdan; Apratim Dey; Shubham Bhardwaj

Gamma-Ray Bursts (GRBs), due to their high luminosities are detected up redshift 10, and thus have the potential be vital cosmological probes of early processes in universe. Fulfilling this requires a large sample GRBs with known redshifts, but observational limitations, only 11\% redshifts ($z$). There been numerous attempts estimate via correlation studies, most which led inaccurate predictions. To overcome this, we estimated GRB an ensemble supervised machine learning model that uses X-ray afterglows long-duration observed by Neil Gehrels Swift Observatory. The strongly correlated (a Pearson coefficient 0.93) root mean square error, namely average squared error $\langle\Delta z^2\rangle$, 0.46 showing reliability method. addition afterglow parameters improves predictions considerably 63\% compared previous results peer-reviewed literature. Finally, use our infer 154 GRBs, increase long plateaus 94\%, significant milestone for enhancing population studies require samples redshift.