Using Multivariate Imputation by Chained Equations to Predict Redshifts of Active Galactic Nuclei

Abstract

Redshift measurement of active galactic nuclei (AGNs) remains a time-consuming and challenging task, as it requires follow up spectroscopic observations and detailed analysis. Hence, there exists an urgent requirement for alternative redshift estimation techniques. The use of machine learning (ML) for this purpose has been growing over the last few years, primarily due to the availability of large-scale galactic surveys. However, due to observational errors, a significant fraction of these data sets often have missing entries, rendering that fraction unusable for ML regression applications. In this study, we demonstrate the performance of an imputation technique called Multivariate Imputation by Chained Equations (MICE), which rectifies the issue of missing data entries by imputing them using the available information in the catalog. We use the Fermi-LAT Fourth Data Release Catalog (4LAC) and impute 24% of the catalog. Subsequently, we follow the methodology described in Dainotti et al. (ApJ, 2021, 920, 118) and create an ML model for estimating the redshift of 4LAC AGNs. We present results which highlight positive impact of MICE imputation technique on the machine learning models performance and obtained redshift estimation accuracy.

Inferring the Redshift of More than 150 GRBs with a Machine-learning Ensemble Model

“…In previous literature (Gibson et al 2022), similar methodologies have been applied to active galactic nuclei data from the fourth Fermi-LAT catalog with no noticeable addition to the uncertainty of the resulting data distribution. In fact, the constructed ML model strictly benefits from its application due to the increased size of the data set.…”

Gamma-Ray Bursts as Distance Indicators by a Statistical Learning Approach

“…We tested the imputed and original distributions with the Kolmogorov-Smirnov test, which yielded a significance level of 0.5. Previously, Gibson et al (2022) applied MICE to impute missing values of active galactic nuclei. There, we demonstrated that this imputation method does not add any additional uncertainty to the variable's original distribution and helps strengthen the final training set used for the statistical learning application by enlarging it.…”