A Local Dwarf Galaxy Search Using Machine Learning

Abstract We present a machine learning search for local, low-mass galaxies ( z < 0.02 and 10 6 M ⊙ < M * < 10 9 M ⊙ ) using combined photometric data from the Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys (hereafter, Legacy Survey) and the Wide-field Infrared Survey Explorer survey. We introduce the spectrally confirmed training sample, discuss evaluation metrics, investigate features, compare different machine learning algorithms, and find that a seven-class neural network classification model is highly effective in separating the signal (local, low-mass galaxies) from various contaminants, reaching a precision of 95% and a recall of 76%. The principal contaminants are nearby sub- L * galaxies at 0.02 < z < 0.05 and nearby massive galaxies at 0.05 < z < 0.2. We find that features encoding surface brightness information are essential to achieving a correct classification. Our final catalog, which we make available, consists of 112,859 local, low-mass galaxy candidates, of which 36,408 have high probability ( p signal > 0.95), covering the entire Legacy Survey DR9 footprint. Using DESI-Early Data Release public spectra and data from the Satellites around Galactic Analogs and Exploration of Local Volume Satellites surveys, we find that our model has a precision of ∼100%, 96%, and 97%, respectively, and a recall of ∼51%, 68%, and 53%, respectively. The results of those independent spectral verifications demonstrate the effectiveness and efficiency of our machine learning classification model.