Performance of morphological classifiers for galaxy mergers compared to current machine learning methods

Non-parametric morphological statistics can be useful in efficiently determining galaxy merger classifications. This work is aimed at comparing the performance of morphological merger classifiers to state-of-the-art machine learning (ML) models. A secondary aim is to produce updated criteria for mergers based on non-parametric morphological statistics. The Gini coefficient (G), M_20 statistic, and concentration (C) were calculated for mock Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) images based on the IllustrisTNG and Horizon-AGN simulations, along with observations from HSC-SSP. IllustrisTNG images were used to find the line that best separates mergers and non-mergers in 2D morphological space with a Markov chain Monte Carlo (MCMC) method. Based on the MCMC results, we classified galaxies with G>(-0.267 +(0.143 20 or G>(0.162 as mergers. These criteria had precisions levels of 69.5% and 72.3% when applied to previously unseen IllustrisTNG mock HSC-SSP images, respectively. The precision of the morphological classifications are consistent with state-of-the-art ML methods. The morphological classifiers were found to be effective at selecting only pre-mergers. On the other hand, post-merger galaxies are indistinguishable from non-mergers in terms of their G, M_20, and C values. Morphological classifiers displayed a similar robustness to new data to ML methods up to a redshift of sim0.52 and maintained the robustness better than ML methods based on convolutional neural networks (CNNs) in the redshift range 0.52<z<1. This work presents updated morphological classifiers with the capacity to achieve a similar level of precision to that of ML-based merger classifiers that have a high robustness with regard to new data. Updated morphological statistics are needed to identify the features of post-merger galaxies.