Predicting the redshift of gamma-ray loud AGNs and GRBs using Supervised machine learning Active galactic nuclei

speaker: Aditya Narendra


(AGNs) are very powerful galaxies characterized by extremely bright emissions coming from their central massive black holes. Knowing the redshifts of AGNs provides us with an opportunity to determine their distance to investigate important astrophysical problems, such as the evolution of the early stars and their formation, along with the structure of early galaxies. However, redshift determination is challenging because it requires detailed follow-up of multiwavelength observations, often involving various astronomical facilities. In this presentation I will discuss about the methodology developed by our team, where we apply a powerful machine learning technique called SuperLearner to estimate the redshift of gamma-ray loud AGNs from the Fermi 4th LAT catalog (4LAC). Using the 4LAC’s observed properties we train the machine learning model on 1112 AGNs, obtaining a correlation of 75% between the predicted and observed redshift. We also explore the application of an imputation method called Multivariate Imputation by Chained Equations (MICE), using which we impute missing data for 24% of the catalog and proceed to investigate its effects on the redshift estimation. Further, we also explore the application of bias corrections and feature engineering for improving our results. Finally, we provide predicted redshift for 300 BL Lacertae Quasars of the 4LAC using our methodology. Similar method is applied to GRBs. I will present the results for GRBs as well.