Summary 

In this chapter, we have learned how graph-based ML techniques can be used to solve many different problems.  

In particular, we have seen that the same algorithm (or a slightly modified version of it) can be adapted to solve apparently very different tasks such as link prediction, community detection, and graph similarity learning. We have also seen that each problem has its own peculiarities, which have been exploited by researchers in order to design more sophisticated solutions.

In the next chapter, we will explore real-life problems that have been solved using ML.