On selecting heuristic function subset for domain independent-planning

• Ano de defesa: 2016
• Autor(a) principal: Zarate, Marvin Abisrror
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Inteligência artificial; Algoritmos; Programação heurística; Heurística; Ciência da Computação
• Link de acesso: http://www.locus.ufv.br/handle/123456789/8541

In this dissertation we present greedy methods for selecting a subset of heuristic functions from a large pool of possibilities with the objective of reducing the running time of search algorithms. Previous works showed that search can be faster if several smaller pattern databases are used instead of one large pattern database. Our methods are able to select good heuristics from a large set of heuristic functions to guide A* search. We implemented our method in Fast Downward and showed empirically that it produces heuristics which outperform the state-of-the-art planners in the International Planning Competition benchmarks.