Identify Main-sequence Binaries from the Chinese Space Station Telescope Survey with Machine Learning. II. Based on Gaia and GALEX

The statistical characteristics of double main-sequence (MS) binaries are essential for investigating star formation, binary evolution, and population synthesis. Our previous study proposed a machine learning-based method to identify MS binaries from MS single stars using mock data from the Chinese Space Station Telescope (CSST). We further utilized detection efficiencies and an empirical mass ratio distribution to estimate the binary fraction within the sample. To further validate the effectiveness of this method, we conducted a more realistic sample simulation, incorporating additional factors such as metallicity, extinction, and photometric errors from CSST simulations. The detection efficiency for binaries with mass ratios between 0.2 and 0.7 reached over 80%. We performed a detailed observational validation using the data selected from the Gaia Sky Survey and Galaxy Evolution Explorer. The detection efficiency for MS binaries in the observed sample was 65%. The binary fraction can be inferred with high precision for a set of observed samples, based on accurate empirical mass ratio distribution.