I guess an interpretable network of (graph) components (or subcomponents) that
uses hypergraphs to refer to them could work in the way I am thinking. To use
them in a variety of what-if scenarios the lines as well as the nodes could be
varied. But I think this would still be in the Relevancy Problem (so to speak).
Maybe the lines do not need to hold values (weights) but then some
associations would have to noted in hypergraph formulas. You want the program
to try different weights of the associative links as well as the nodes. But
this is hyper-complex (if you'll excuse the pun).
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/T1675e049c274c867-M12905e4b4b97fb76d07d5d55
Delivery options: https://agi.topicbox.com/groups/agi/subscription
