首页 | 本学科首页   官方微博 | 高级检索  
     


A Multilevel Probabilistic Beam Search Algorithm for the Shortest Common Supersequence Problem
Authors:José E. Gallardo
Affiliation:Departamento de Lenguajes y Ciencias de la Computación, Universidad de Málaga, Málaga, Spain.; Universitat Rovira i Virgili, Spain,
Abstract:The shortest common supersequence problem is a classical problem with many applications in different fields such as planning, Artificial Intelligence and especially in Bioinformatics. Due to its NP-hardness, we can not expect to efficiently solve this problem using conventional exact techniques. This paper presents a heuristic to tackle this problem based on the use at different levels of a probabilistic variant of a classical heuristic known as Beam Search. The proposed algorithm is empirically analysed and compared to current approaches in the literature. Experiments show that it provides better quality solutions in a reasonable time for medium and large instances of the problem. For very large instances, our heuristic also provides better solutions, but required execution times may increase considerably.
Keywords:
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号