Fig. 8

The GDT-TS score distribution of the training dataset. This mixture distribution is approximately composed of two Gaussian distributions \({\mathcal{N}}_{1} \left( {0.241, 0.015} \right)\) and \({\mathcal{N}}_{2} \left( {0.667, 0.018} \right)\). The red line represents the mean of \({\mathcal{N}}_{1}\) and the blue line the mean of \({\mathcal{N}}_{2}\)