I'm having trouble finding information on this because I am not familiar with the nomenclature for graph theory. If you don't know of a specific algorithm, or if you know no such algorithm exists, giving me the terms to search for in the literature would be helpful. If possible I would prefer an algorithm was at its core bayesian, using what has been seen before as a standard for what it can expect to encounter in the future. How does the answer to my question change if we have some information about the graph, like the number of nodes?
A secondary question would be to predict the labeling of nodes and edges given the nodes that have already seen. So if the graph is a simple chain, and the first node is "A" and the second node is "B" we would naturally predict the next node would be labelled "C".
Dear Julio,
I suggest you to see links and attached file on topic.
https://www.cs.upc.edu/~dariog/PhD-Thesis-Link-Prediction-DGG.pdf
Article Intelligent route planning system based on interval computing
Article A distributed algorithm to maintain and repair the trail net...
Article Small-World Brain Networks Revisited
Best regards
Sometimes I think about the idea of complexity theory with an adversary. An algorithm realised as Turing machine is intended to make propositions about the graph. An adversary may before each state change inspect the TM, its state, and already written output. While the TM reads parts of the graph as input, the opponent of the algorithm is allowed to change the hitherto unread graph that lies ahead. He wins if in all cases except for a finite number he is able to spoil the TM's author's proposition.
The adversary can be an algorithm as well.
Regards,
Joachim
You may find useful my old "A useful Transform of Standard Input Data
for a Classical NP-complete problem" European Journal of Operational Research 19(1985) 390-396, North Holland
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs