Modularity and Silhouette Scores
The modularity of a network gives us an idea about the internal structure of the network. When the structure is highly modularized, it means the network essentially consists of some loosely coupled components (sub-networks) and its modularity is moving to its highest end of 1. In contrast, if all the components of the network are tightly coupled, then its modularity is moving towards the low end of the metric of 0.
The silhouette score of a cluster measures the homogeneity of the cluster, which is to answer the question whether cluster members are lumped together based on what they have in common in some aspects. In other words, a cluster with a high silhouette score is considered more meaningful than a grouping with a low silhouette score.
References:
The silhouette score of a cluster measures the homogeneity of the cluster, which is to answer the question whether cluster members are lumped together based on what they have in common in some aspects. In other words, a cluster with a high silhouette score is considered more meaningful than a grouping with a low silhouette score.
References:
- Newman, M.E.J. (2006). Modularity and community structure in networks.PNAS, 103(23), 8577—8582.
- Rousseeuw, P.J. (1987). Silhouettes: A graphical aid to the interpretationand validation of cluster analysis. Journal of Computational and AppliedMathematics, 20, 53—65.