Wrapping Up
In Chapter 2, you looked at k-means clustering. Using the same data in this chapter, you tackled network graphs and clustering via modularity maximization. You should feel pretty good about your data mining chops by now. In more detail, here are some items you learned:
- How network graphs are visually represented as well as how they're represented numerically using adjacency and affinity matrices
- How to load a network graph into Gephi to augment Excel's visualization deficiencies
- How to prune edges from network graphs via the r-neighborhood graph. You also learned the concept of a kNN graph, which I recommend you go back and tinker with.
- The definitions of node degree and graph modularity and how to calculate modularity scores for grouping two nodes together
- How to maximize graph modularity using a linear optimization model and divisive clustering
- How to maximize graph modularity in Gephi and export the results
Now, you may be wondering, “John, why in the...