Fig. 8From: Machine learning-assisted high-throughput exploration of interface energy space in multi-phase-field model with CALPHAD potentialLeft) Example clusters identified by visual inspection of the data, (i) invariant, ii) wetting, iii) dewetting, and v) droplet. Right) VAE latent space representation of microstructure. The classification boxes indicated by dashed lines are visually set to roughly identify the regions in kPCA space where each microstructure class belongsBack to article page