Fine-Grained Complexity of Constraint Satisfaction Problems through Partial Polymorphisms: A Survey (Dedicated to the memory of Professor Ivo Rosenberg) - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2019

Fine-Grained Complexity of Constraint Satisfaction Problems through Partial Polymorphisms: A Survey (Dedicated to the memory of Professor Ivo Rosenberg)

Abstract

Constraint satisfaction problems (CSPs) are combi-natorial problems with strong ties to universal algebra and clone theory. The recently proved CSP dichotomy theorem states that finite-domain CSPs are always either tractable or NP-complete. However, among the intractable cases there is a seemingly large variance in complexity, which cannot be explained by the classical algebraic approach using polymorphisms. In this contribution we will survey an alternative approach based on partial polymorphisms, which is useful for studying the fine-grained complexity of NP-complete CSPs. Moreover, we will state some challenging open problems in the research field.
Fichier principal
Vignette du fichier
survey_paper_ismvl-REV.pdf (293.94 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02190089 , version 1 (21-07-2019)

Identifiers

  • HAL Id : hal-02190089 , version 1

Cite

Miguel Couceiro, Lucien Haddad, Victor Lagerkvist. Fine-Grained Complexity of Constraint Satisfaction Problems through Partial Polymorphisms: A Survey (Dedicated to the memory of Professor Ivo Rosenberg). ISMVL2019 - IEEE 49th International Symposium on Multiple-Valued Logic, May 2019, Fredericton, NB, Canada. ⟨hal-02190089⟩
64 View
70 Download

Share

Gmail Facebook X LinkedIn More