We investigate the complexity of counting problems that belong to the complexity class #P and have an easy decision version. These problems constitute the class #PE which has some well-known representatives such as #PERFECT MATCHINGS, #DNF-SAT, and NONNEGATIVE PERMANENT. An important property of these problems is that they are all #P-complete, in the Cook sense, while they cannot be #P-complete in the Karp sense unless P = NP. We study these problems in respect to the complexity class TotP, which contains functions that count the number of all paths of a PNTM. We first compare TotP to #P and #PE and show that FP ⊆ TotP ⊆ #PE ⊆ #P and that the inclusions are proper unless P = NP. We then show that several natural #PE problems -including the ones mentioned above -belong to TotP. Moreover, we prove that TotP is exactly the Karp closure of self-reducible functions of #PE. Therefore, all these problems share a remarkable structural property: for each of them there exists a polynomial-time nondeterministic Turing machine which has as many computation paths as the output value. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Pagourtzis, A., & Zachos, S. (2006). The complexity of counting functions with easy decision version. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4162 LNCS, pp. 741–752). Springer Verlag. https://doi.org/10.1007/11821069_64
Mendeley helps you to discover research relevant for your work.