Hawker & Integer
I’ve been thinking about how a simple shortest‑path algorithm could solve a strategic resource‑allocation problem. Have you tackled something like that before?
Yes, I’ve tried to model a resource‑allocation problem as a shortest‑path instance. I map each resource state to a node, each allocation action to an edge, and then run Dijkstra or Bellman‑Ford to find the minimal cost sequence. The trick is making sure the graph stays small enough so the algorithm runs in real time; otherwise the state explosion defeats the purpose. It works well when the constraints are linear, but once you add non‑linear interactions the graph becomes too dense. Still, it’s a neat exercise in reducing a combinatorial problem to a well‑understood algorithmic framework.
Sounds solid. Keep the state space trimmed by pruning infeasible branches early. If non‑linear terms pop up, try linearizing or bounding them so the graph stays sparse. That way you maintain real‑time performance while still using the shortest‑path core.
That’s a good plan—prune before you get stuck, and linearize only where it really matters. Keeps the graph manageable and the run time predictable.