10.10.2 Properties of the input data

The initial (super-set) of candidates was identified by supervised classification (Section 10.3). The training set for the classifier consisted of light curves of previously known exoplanets, with transits that were clearly visible in the Gaia data. Each light curve contained all three photometric bands, $G$, $G_{\mathrm{RP}}$, and $G_{\mathrm{BP}}$.