I'm building the network to validate the directional graph structure.
So I have matrices to reflect the structure and edge properties.
On the end I'm expecting to have a network works as a heuristic function to roughly evaluate the graph to say which one is "better".
I wanted to have all information in one file line, read it by CSV Reader, create a schema and pull to network (train, experiment with and so on).
I can have in one line more then 1000 cells which need to proceed be the network after all.
Can somebody validate my thoughts or give some feedback?
Maybe somebody did similar task or has better overview how to model this.