Showing posts with label blind. Show all posts
Showing posts with label blind. Show all posts

Sunday, 18 January 2009

Classification of Strategies for State Space Search

Blind (exhaustive, uninformed, naive) – explore the state space systematically without any “suspicion” where the final state might be and direction of the search to it.
Breadth-first search;
Depth-first search (depth-first search with leap-frogging (or just depth-first search), depth-first search with backtracking (or just Backtracking)
Uniform-cost search;
Iterative deepening;
Etc.
Heuristic (informed)- exploit state descriptions to select the “most promising” node

Blind State Space Search : General Idea

We will be looking here for one solution only

Expand a state (find its successors / descendants / children / adjacent states)
Forward strategy
Backwards strategy
Bidirectional search

Exit:
A final state has been reached [Success]
All reachable states have been explored and no solution has been found [Failure]

After success – usually reconstruct the solution path