What algorithm paradigms do DFS and BFS search belong to? Are search algorithms different from dp or greedy or have some of their characteristics?EG Best first search(greedy)(making locally optimal choices) https://stackoverflow.com/questions/8374308/is-greedy-best-first-search-different-from-best-first-search
The following observations are only valid if you use DFS/BFS to find an element in a graph. DFS/BFS is also often used for different purposes. E.g. when you want to explore the complete graph, either in a depth first or breath first manner.
Both algorithms fall in the category “Search”.
I think also the paradigm “dynamic programming” applies to both. Because in both algorithms you mark nodes as “visited”, so that you doesn’t search at nodes, which you searched before.
DFS can also be seen as a “greedy” algorithm. At least if you look for a path to an element, it will report the first one that it finds, and the length of the path might not be optimal. BFS on the other side is not greedy. But “bruteforce” might apply here.
Generally I find paradigms a bit fuzzy and they leave room for a lot of interpretation. So the answer is not 100% correct. Depending on what you define as paradigm, the answer will vary a bit.