Anyone could explain to me what is the difference between graph convolutional network (GCN) and random walk? or they are the same?

Any further explanation will be much appreciated.


1 Answer 1


Graph convolutional networks are generalizations of standard convolutional networks to irregular graphs. Convolutional networks operate on regular n-dimensional matrices/tensors.

Random walk is a whole another concept. You have a starting point, and randomly choose where to among your neighbours via some set of defined probability rules. In standard 2D, it's like going towards south/north/east/west randomly some number of steps. Random walks can also be performed over graphs, since at each step, you'll choose a neighbour and iterate.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.