Why did Deep First Search say it suffers from endless cycles?

Question

I have read about DFS and BFS many times, but I have this doubt that has been holding my mind back since ancient times. Many articles mention that DFS can get stuck in endless loops.

As far as I know, this restriction can be easily removed by tracking visited sites. In fact, in all the books I read, this little check is part of DFS.

So why are "infinite loops" referred to as a DFS flaw? Is it just because the original DFS algorithm did not have this check for visited nodes? Please explain.

Answer 1

(1) In graph search algorithms [often used for AI], the main advantage of DFS is space efficiency . This is his main advantage in BFS. However , if you track visited sites, you lose this advantage , since you need to store all visited sites in memory. Do not forget that the size of the visited nodes increases sharply over time, and for very large / endless graphs - it may not fit in memory.

(2) Sometimes DFS can be in an infinite branch [in infinite graphs]. An infinite branch is a branch that does not end [always has “more sons”], and also does not fall into your target node, so for DFS you can continue to expand this branch endlessly and “skip” the good branch that leads to the target node.

Bonus:
You can overcome this drawback in DFS by saving a relatively small amount of memory using a combination of DFS and BFS: Iterative deepening DFS

Answer 2

A regular DFS algorithm tracks nodes. The local search algorithm does not track conditions and behaves with amnesia. Therefore, I think that the loop mainly refers to an infinite branch (branches with infinite possible states). In this case, DFS simply drops and becomes too focused on one branch.