Automated, adaptive, and optimized search for CSPs via cuckoo search

Ricardo Soto, Broderick Crawford, Javier Flores, Felipe Mella, Cristian Galleguillos, Franklin Johnson, Fernando Paredes

Resultado de la investigación: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

Constraint Programing is a programming paradigm devoted to the efficient solving of constraint satisfaction problems (CSPs). A CSP is a formal problem representation mainly composed of variables and constraints defining relations among those variables. The resolution process of CSPs is commonly carried out by building and exploring a search tree that holds the possibles solutions. Such a tree is dynamically created by interleaving two different phases: enumeration and propagation. During enumeration, the variables and values are chosen to build the possible solution, while propagation intend to delete the values having no chance to reach a feasible result. Autonomous Search is a new technique that gives the ability to the resolution process to be adaptive by re-configuring its enumeration strategy when poor performances are detected. This technique has exhibited impressive results during the last years. However, such a re-configuration is hard to achieve as parameters are problem-dependent and their best configuration is not stable along the search. In this paper, we introduce an Autonomous Search framework that incorporates a new optimizer based on Cuckoo Search able to efficiently support the re-configuration phase. Our goal is to provide an automated, adaptive, and optimized search system for CSPs. We report encouraging results where our approach clearly improves the performance of previously reported Autonomous Search approaches for CSPs.

Idioma originalInglés
Título de la publicación alojadaAdvances in Swarm and Computational Intelligence - 6th International Conference, ICSI 2015 held in conjunction with the 2nd BRICS Congress, CCI 2015, Proceedings
EditoresAlexander Gelbukh, Ying Tan, Swagatam Das, Andries Engelbrecht, Fernando Buarque, Yuhui Shi
EditorialSpringer Verlag
Páginas436-447
Número de páginas12
ISBN (versión impresa)9783319204659
DOI
EstadoPublicada - 2015
Publicado de forma externa
Evento6th International Conference on Swarm Intelligence, ICSI 2015 held in conjunction with the 2nd BRICS Congress on Computational Intelligence, CCI 2015 - Beijing, China
Duración: 25 jun. 201528 jun. 2015

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen9140
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia6th International Conference on Swarm Intelligence, ICSI 2015 held in conjunction with the 2nd BRICS Congress on Computational Intelligence, CCI 2015
País/TerritorioChina
CiudadBeijing
Período25/06/1528/06/15

Huella

Profundice en los temas de investigación de 'Automated, adaptive, and optimized search for CSPs via cuckoo search'. En conjunto forman una huella única.

Citar esto