Questions tagged [graph-search]

For questions related to the graph search. As opposed to a tree search, a graph search uses a list or set, called the closed list (or explored set), which contains the already visited and expanded nodes (or states) of the search space (which is usually represented as a graph, both in the case of tree and graph searches), so that not to revisit these already visited nodes.

Filter by
Sorted by
Tagged with
1 vote
1 answer

GREED - preservation theoretical properties in the GED(graph edit distance) pridiction

In this paper "GREED: A Neural Framework for Learning Graph Distance Functions", function F is defined to satisfy metric property and triangle inequality property. I wonder how can I prove ...
19 votes
1 answer

What is the difference between tree search and graph search?

I have read various answers to this question at different places, but I am still missing something. What I have understood is that a graph search holds a closed list, with all expanded nodes, so ...
2 votes
0 answers

Why do we use the tree-search version of breadth-first search or A*?

In Artificial Intelligence A Modern Approach, search algorithms are divided into tree-search version and graph-search version, where the graph-search version keeps an extra explored set to avoid ...
1 vote
0 answers

What is a example showing that the tree-based variant for the greedy best-first search is incomplete?

I understand that a tree-based variant will have nodes repeatedly added to the frontier. How do I craft an example where a particular goal node is never found. Is this example valid. On the other ...
2 votes
1 answer

In the graph search version of A*, can I stop the search the first time I encounter the goal node?

I am going through Russel and Norvig's Artificial Intelligence: A Modern Approach (3rd edition). I was reading the part regarding the A* algorithm A* graph search version is optimal when heuristic ...