The Single Frontier Bi-Directional Search (SBS) framework was recently introduced. A node in SBS corresponds to a pair of states, one from each of the frontiers and it uses front-tofront heuristics. In this paper we present an enhanced version of SBS, called eSBS, where pruning and caching techniques are applied, which significantly reduce both time and memory needs of SBS. We then present a hybrid of eSBS and IDA* which potentially uses only the square root of the memory required by A* but enables to prune many nodes that IDA* would generate. Experimental results show the benefit of our new approaches on a number of domains. Copyright © 2012, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
Efficient single frontier bidirectional search / Lippi, Marco; Ernandes, Marco; Felner, Ariel. - (2012), pp. 49-56. (Intervento presentato al convegno 5th International Symposium on Combinatorial Search, SoCS 2012 tenutosi a Niagara Falls, ON, can nel 2012).
Efficient single frontier bidirectional search
LIPPI, MARCO;
2012
Abstract
The Single Frontier Bi-Directional Search (SBS) framework was recently introduced. A node in SBS corresponds to a pair of states, one from each of the frontiers and it uses front-tofront heuristics. In this paper we present an enhanced version of SBS, called eSBS, where pruning and caching techniques are applied, which significantly reduce both time and memory needs of SBS. We then present a hybrid of eSBS and IDA* which potentially uses only the square root of the memory required by A* but enables to prune many nodes that IDA* would generate. Experimental results show the benefit of our new approaches on a number of domains. Copyright © 2012, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
| File | Dimensione | Formato | |
|---|---|---|---|
|
sfbdsh.pdf
Accesso riservato
Tipologia:
AAM - Versione dell'autore revisionata e accettata per la pubblicazione
Dimensione
251.78 kB
Formato
Adobe PDF
|
251.78 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris
