Approximation via Correlation Decay when Strong Spatial Mixing Fails
- Submitting institution
-
University of Oxford
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2075
- Type
- D - Journal article
- DOI
-
10.1137/16M1083906
- Title of journal
- SIAM Journal on Computing
- Article number
- -
- First page
- 279
- Volume
- 48
- Issue
- 2
- ISSN
- 1095-7111
- Open access status
- Compliant
- Month of publication
- March
- Year of publication
- 2019
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
4
- Research group(s)
-
-
- Citation count
- 3
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This SICOMP paper extends a conference version in ICALP�16. Correlation decay is a key algorithmic technique that is important in understanding the limits of computational tractability in approximate counting. This paper develops a new amortisation method to account for the success of correlation decay algorithms. It overcomes the worst-case barriers that obstructed previous analyses . The new method gave an efficient algorithm for approximately counting satisfying assignments in monotone CNF formulas, a canonical problem where it was open whether and how correlation decay can be applied effectively (described as a �major open question� by Lu, Yang, and Zhang STACS�16).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -