Driver Attribute Filling for Genes in Interaction Network via Modularity Subspace-Based Concept Learning from Small Samples

Figure 3

Performance evaluation of the compared methods on iRefIndex network via ROC curves and AUC values. (a) ROC curve for prostate cancer. (b) ROC curve for thyroid cancer. (c) AUC for prostate cancer. (d) AUC values for thyroid cancer.

(a)

(b)

(c)

(d)