Abstract
The study of fitness landscapes is important for increasing our understanding of local-search based heuristics and evolutionary algorithms. The number of acceptable solutions in the landscape is a crucial factor in measuring the difficulty of combinatorial optimization and decision problems. This paper estimates this number from statistics on the number of repetitions in the sample history of a search. The approach is applied to the problem of counting the number of satisfying solutions in random and structured SAT instances. © Springer-Verlag 2004.
Cite
CITATION STYLE
Reeves, C. R., & Aupetit-Bélaidouni, M. (2004). Estimating the number of solutions for SAT problems. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3242, 101–110. https://doi.org/10.1007/978-3-540-30217-9_11
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
