Reply to thread

Message: [QUOTE="Joshua Cranmer, post: 3597165"] Keeping score, that change should make it O(N^4): for each starting node for each node in path (actually N/2, but ignoring constants) for each possible next node for T_1 for each possible previous node for T_2 Forgive me for asking, but why multiply the two costs instead of adding them together? Answers my question in other post, but still a disconcerting answer, since (as I explained again other there) you lose something by moving to a node, unlike other working greedy algorithms where "moving" to a node is only expanding a search tree. Taking this time to expound on what I think the flaw in the algorithm is: you're ignoring the opportunity cost of making a move. [/QUOTE]