Figure 3
From: Statistical Modeling of the Default Mode Brain Network Reveals a Segregated Highway Structure
Plots illustrating both edgewise and structural fit. TOP: Observed network (middle) versus the conditional edgewise network predicted from the model with only edges, hemisphere, and distance (left) or from all features (right). From these, we can see that the conditional edgewise prediction is producing networks that closely map the observed networks. Further, the model that accounts for the topological features (right) produces more similar structure than the model which does not (left). BOTTOM: Goodness of fit for the model with only hemisphere and distance (left) versus all features (right). Boxplots reflect values from simulation based on the fitted cGERGM, with the red bars representing the mean of the simulations whereas the blue bars represent the values from the observed network. The curve to the right of each boxplot represents simulated (red) and observed (blue) degree distributions. The left model is specified to include only the edges term and the exogenous predictors of hemisphere and distance. The right model includes all topological and exogenous effects. While the model without topological features fits quite poorly, when modeling topological features our models fit quite well.
