Fig. 4: Validation of trained models.
From: Machine learning-enabled chemical space exploration of all-inorganic perovskites for photovoltaics
Parity plot between crystal graph convolution neural network (CGCNN)-predicted and DFT (PBESol)-calculated (a) ∆Hdecomp and b bandgap on the test set. c: Confusion matrix of classification test. “Non-indirect” includes direct, metallic, and semi-metallic materials.
