Jump to content

Beam stack search

From Wikipedia, the free encyclopedia
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Beam stack search[1] is a search algorithm that combines chronological backtracking (that is, depth-first search) with beam search and is similar to depth-first beam search.[2] Both search algorithms are anytime algorithms that find good but likely sub-optimal solutions quickly, like beam search, then backtrack and continue to find improved solutions until convergence to an optimal solution.

Implementation

Beam stack search uses the beam stack as a data structure to integrate chronological backtracking with beam search and can be combined with the divide and conquer algorithm technique, resulting in divide-and-conquer beam-stack search.

Alternatives

Beam search using limited discrepancy backtracking[2] (BULB) is a search algorithm that combines limited discrepancy search with beam search and thus performs non-chronological backtracking, which often outperforms the chronological backtracking done by beam stack search and depth-first beam search.

References

  1. ^ Zhou, Rong; Hansen, Eric A. (2005). "Beam-Stack Search: Integrating Backtracking with Beam Search" (PDF). CiteSeerx10.1.1.71.4147.
  2. ^ a b Furcy, David. Koenig, Sven. "Limited Discrepancy Beam Search". 2005. "Archived copy" (PDF). Retrieved 2007-12-22.