In several applications, the data
can be naturally represented by graph
structures. The simplest kind of graph
structures are single nodes, and sequences,
but in many application domains, the
information is organized in more complex
graph structures such as trees, acyclic
graphs, or cyclic graphs.
In machine learning, the structured
data is often associated with the
goal of either supervised or unsupervised
learning from examples such that a
function maps a graph G and one of
its nodes n to a vector of reals.
In those applications, the goal consists
of learning from examples a function
that maps a graph G and one of its
nodes n to a vector of reals:
