Dynamic Potential Search — A New Bounded Suboptimal Search

Gilon, Daniel; Felner, Ariel; Stern, Roni

doi:10.1609/socs.v7i1.18392

Cited by 16 publications

(20 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For all experiments, we use a 30-minute timeout. Our algorithms are compared with Focal Search (FS) using the same neural-net heuristic; and two other state-of-the-art boundedsuboptimal search algorithms: Weighted A* (WA*) and Dynamic Potential Search (DPS) (Gilon, Felner, and Stern 2016).…”

Section: Empirical Evaluationmentioning

confidence: 99%

K-Focal Search for Slow Learned Heuristics (Extended Abstract)

Greco

Toro

Hernández-Ulloa

et al. 2022

SOCS

View full text Add to dashboard Cite

Learned heuristics, though inadmissible, can provide very good guidance for bounded-suboptimal search. Given a single search state s and a learned heuristic h, evaluating h(s) is typically very slow relative to expansion time, since state-of-the-art learned heuristics are implemented as neural networks. However, by using a Graphics Processing Unit (GPU), it is possible to compute heuristics using batched computation. Existing approaches to batched heuristic computation are specific to satisficing search and have not studied the problem in the context of bounded-suboptimal search. In this paper, we present K-Focal Search, a bounded suboptimal search algorithm that in each iteration expands K nodes from the FOCAL list and computes the learned heuristic values of the successors using a GPU. We experiment over the Rubik's Cube domain using DeepCubeA, a very effective inadmissible learned heuristic. Our results show that K-Focal Search benefits both from batched computation and from the diversity in the search introduced by its expansion strategy. Over standard FS, it improves runtime by a factor of 6, expansions by up to three orders of magnitude, and finds better solutions, keeping the theoretical guarantees of Focal Search.

show abstract

Section: Empirical Evaluationmentioning

confidence: 99%

K-Focal Search for Slow Learned Heuristics (Extended Abstract)

Greco

Toro

Hernández-Ulloa

et al. 2022

SOCS

View full text Add to dashboard Cite

show abstract

“…However, such an algorithm might expand nodes n with f (n) > C * whose g(n) ≤ C * . Gilon, Felner, and Stern (2016) denoted algorithms as reasonable if they have a best-first structure (i.e. an open list and an expansion rule), and they prune any node n with f (n) > C, where C an upper bound on the cost.…”

Section: The Reasonableness Propertymentioning

confidence: 99%

“…Note that since f (n) ≤ lb(n) and C * ≤ C, the redefinition of the reasonable property is tighter than the original definition of Gilon, Felner, and Stern (2016). Theorem 2.…”

Section: The Reasonableness Propertymentioning

confidence: 99%

Improving Bidirectional Heuristic Search by Bounds Propagation

Shperberg

Felner

Shimony

et al. 2021

SOCS

View full text Add to dashboard Cite

Recent work in bidirectional heuristic search characterize pairs of nodes from which at least one node must be expanded in order to ensure optimality of solutions. We use these findings to propose a method for improving existing heuristics by propagating lower bounds between the forward and backward frontiers. We then define a number of desirable properties for bidirectional heuristic search algorithms, and show that applying the bound propagations adds these properties to many existing algorithms (e.g. to the MM family of algorithms). Finally, experimental results show that applying these propagations significantly reduce the running time of various algorithms.

show abstract

“…Dynamic Potential Search (DPS) (Gilon, Felner, and Stern 2016) uses a dynamic priority function which can be implemented using a single priority queue. But, because of the dynamic nature of the search, DPS must occasionally re-order the states on the priority queue.…”

Section: Background and Related Workmentioning

confidence: 99%

“…These domains are pathfinding on gridbased benchmarks (Sturtevant 2012) with an octile heuristic, the sliding-tile puzzle with Manhattan distance, the heavy sliding-tile puzzle with heavy Manhattan distance, and the heavy pancake puzzle with a modified GAP heuristic. The heavy sliding-tile puzzle uses linear weights -the cost of moving tile n is n. The heavy pancake puzzle uses the maximum size of the top and bottom pancake in a stack being flipped as the action cost (Gilon, Felner, and Stern 2016).…”

Section: Study Of Parametersmentioning

confidence: 99%

Revisiting Suboptimal Search

Chen

Sturtevant

Doyle

et al. 2021

SOCS

View full text Add to dashboard Cite

Suboptimal search algorithms can often solve much larger problems than optimal search algorithms, and thus have broad practical use. This paper returns to early algorithms like WA*, A*_e and Optimistic search. It studies the commonalities between these approaches in order to build a new bounded-suboptimal algorithm. Combined with recent research on avoiding node re-expansions in bounded-optimal search, a new solution quality bound is developed, which often provides proof of the solution bound much earlier during the search. Put together, these ideas provide a new state-of-the-art in bounded-optimal search.

show abstract

Dynamic Potential Search — A New Bounded Suboptimal Search

Cited by 16 publications

References 12 publications

K-Focal Search for Slow Learned Heuristics (Extended Abstract)

K-Focal Search for Slow Learned Heuristics (Extended Abstract)

Improving Bidirectional Heuristic Search by Bounds Propagation

Revisiting Suboptimal Search

Contact Info

Product

Resources

About