CHARACTERISTIC SKY BACKGROUND FEATURES AROUND GALAXY MERGERS

Abstract

In the context of finding galaxy merger in large-scale surveys, we applied Machine
Learning algorithms that, instead of using the images as it is the current
standard, made used of flux measurements. Training multiple NNs using a
class-balanced dataset of mergers and non-mergers Sloan Digital Sky Survey,
we found that the sky background error parameters could provide a validation
92.64 ± 0.15 % accuracy of and a training accuracy of 92.36 ± 0.21 %.
Moreover, analysing the NN identifications led us to find that a simple decision
diagram using the sky error for two flux filters is enough to get a 91.59 % accuracy.
By understanding how the galaxies vary along the diagram, and trying to
parametrize the methodology in the deeper images of the Hyper Suprime-Cam,
we are currently trying to define and generalize this sky error-based methodology.