Fig. 4

We consider the MS model with the background (resp., template) channels independent and marginally distributed as ER(n, 1/2) (resp., ER(m, 1/2)). Here the template is correlated edge-wise to the induced subgraph of the first m vertices in the background. Here we fix \(m=100\) and we consider n ranging over \(\{100,500,1000,2000\}\). We fix \(\rho = 0.5\) and we apply naive padding (left panel) and centered padding (right panel) before matching the template to the background via M-FAQ with \(s=10\) seeds (averaged over 100 Monte Carlo replicates). In red (resp., green, blue, purple) we plot the results for \(n=100\) (resp., \(n=500\), \(n=1000\), \(n=2000\)). The partially transparent points visualize the accuracy distribution and correspond to individual Monte Carlo replicates