Figure 4.

Figure plots the average detection accuracies on all the data centers when inter-graph distances are calculated using different subsets of node attributes. The classification is performed on the male and female groups of subjects separately to achieve the reported results on (A) training data sets and (B) test data sets.