Fig. 7From: CRPGCN: predicting circRNA-disease associations using graph convolutional network based on heterogeneous networkComparison of the three algorithm modes of 4 different datasets under GRPGCN. a CRPGCN is the method proposed in this paper, which undergoes RWR for similarity aggregation and PCA feature extraction. The CRPGCN-I method is a direct fusion for feature extraction by PCA algorithm after comprehensive similarity calculation; CRPGCN-II is a direct fusion after comprehensive similarity calculation, without RWR for similarity aggregation and PCA feature extraction. b Represents the AUC values of twofold CV, fivefold CV and tenfold CV of the three models under four datasetsBack to article page