Fig. 3
From: A machine learning approach to model solute grain boundary segregation
GB solute concentration by several models. The WC prediction (Eq. (1)) for the Co concentration (solid blue) for the aggregated DOS (see text) using a bulk concentration of 0.2 at%. The LM prediction (DOS replaced by δ-function) using the calculated mean (solid dark red) or most attractive (solid light red) segregation energy. The dotted red line shows the LM model with the effective interaction energy and saturation limit used as fitting parameters. The dotted blue line shows a fit of the WC model based on a Gaussian DOS (see text). The inset shows the corresponding segregation energy DOS and uses the same colour and line scheme
