On the Relative Power of Linear Algebraic Approximations of Graph Isomorphism

Anuj Dawar, Danny Vagnozzi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We compare the capabilities of two approaches to approximating graph isomorphism using linear algebraic methods: the invertible map tests (introduced by Dawar and Holm) and proof systems with algebraic rules, namely polynomial calculus, monomial calculus and Nullstellensatz calculus. In the case of fields of characteristic zero, these variants are all essentially equivalent to the Weisfeiler-Leman algorithms. In positive characteristic we show that the distinguishing power of the monomial calculus is no greater than the invertible map method by simulating the former in a fixed-point logic with solvability operators. In turn, we show that the distinctions made by this logic can be implemented in the Nullstellensatz calculus.
Original languageEnglish
Title of host publication46th International Symposium on Mathematical Foundations of Computer Science (MFCS 2021)
EditorsFilippo Bonchi, Simon J. Puglisi
Place of PublicationDagstuhl, Germany
PublisherSchloss Dagstuhl - Leibniz-Zentrum für Informatik
Number of pages16
Volume202
ISBN (Print)978-3-95977-201-3
DOIs
Publication statusPublished - 18 Aug 2021
Event46th International Symposium on Mathematical Foundations of Computer Science
- Tallinn, Estonia
Duration: 23 Aug 202127 Aug 2021
Conference number: 46

Publication series

NameLeibniz International Proceedings in Informatics (LIPIcs)
PublisherSchloss Dagstuhl -- Leibniz-Zentrum für Informatik
ISSN (Print)1868-8969

Conference

Conference46th International Symposium on Mathematical Foundations of Computer Science
Abbreviated titleMFCS 2021
Country/TerritoryEstonia
CityTallinn
Period23/08/2127/08/21

Keywords / Materials (for Non-textual outputs)

  • Graph isomorphism
  • proof complexity
  • invertible map tests

Fingerprint

Dive into the research topics of 'On the Relative Power of Linear Algebraic Approximations of Graph Isomorphism'. Together they form a unique fingerprint.

Cite this