Hands-On Machine Learning with C++

Newman modularity-based graph clustering algorithm with Dlib

Implementation of this algorithm is based on the work Modularity and community structure in networks by M. E. J. Newman. This algorithm is based on the modularity matrix for a network or a graph and it is not based on particular graph theory, but it has instead some similarities with spectral clustering because it also uses eigenvectors.

The Dlib library implements this algorithm in the newman_cluster() function, which takes a vector of weighted graph edges and outputs the container with cluster indices for each vertex. The initial step for using this algorithm is the definition of graph edges. In the following code sample, we make edges between almost every pair of dataset objects. ...