Machine Learning for Networks

I have made up a workflow for Machine Learning in Networks. My objective is to track disease flow from one node to another node, nodes being persons, and recognize new tracks when new network is feeded. I was able to build a workflow only upto certain extent. I don't know on how to proceed further. I have been through many websites and Knime Forum, but I didn't find any suitable answers. In this networks, I thought of including Python based nodes, so that it simplifies my work. Can anyone please help on how to proceed further?