Performance of Sequential Local Algorithms for the Random NAE-$K$-SAT Problem
Name
Performance of sequential.pdf
Size
355.04 KB
Format
Adobe PDF
Checksum (MD5)
eac7f1d9dc6dc5971577a19c97fb7f5d
Author(s)
Gamarnik, David
Sudan, Madhu
Date Issued
March 2017
Journal
SIAM Journal on Computing
Publisher
Society for Industrial and Applied Mathematics
Citation
Gamarnik, David, and Madhu Sudan. “Performance of Sequential Local Algorithms for the Random NAE-$K$-SAT Problem.” SIAM Journal on Computing 46, no. 2 (January 2017): 590–619.
Version
Final published version
Abstract
We formalize the class of “sequential local algorithms" and show that these algorithms fail to find satisfying assignments on random instances of the “Not-All-Equal-$K$-SAT” (NAE-$K$-SAT) problem if the number of message passing iterations is bounded by a function moderately growing in the number of variables and if the clause-to-variable ratio is above $(1+o_K(1)){2^{K-1}\over K}\ln^2 K$ for sufficiently large $K$. Sequential local algorithms are those that iteratively set variables based on some local information and/or local randomness and then recurse on the reduced instance. Our model captures some weak abstractions of natural algorithms such as Survey Propagation (SP)-guided as well as Belief Propagation (BP)-guided decimation algorithms---two widely studied message-passing--based algorithms---when the number of message-passing rounds in these algorithms is restricted to be growing only moderately with the number of variables. The approach underlying our paper is based on an intricate geometry of the solution space of a random NAE-$K$-SAT problem. We show that above the $(1+o_K(1)){2^{K-1}\over K}\ln^2 K$ threshold, the overlap structure of $m$-tuples of nearly (in an appropriate sense) satisfying assignments exhibit a certain behavior expressed in the form of some constraints on pairwise distances between the $m$ assignments for appropriately chosen positive integer $m$. We further show that if a sequential local algorithm succeeds in finding a satisfying assignment with probability bounded away from zero, then one can construct an $m$-tuple of solutions violating these constraints, thus leading to a contradiction. Along with [D. Gamarnik and M. Sudan, Ann. Probab., to appear], where a similar approach was used in a (somewhat simpler) setting of nonsequential local algorithms, this result is the first work that directly links the overlap property of random constraint satisfaction problems to the computational hardness of finding satisfying assignments.
MIT Department
Sloan School of Management
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1137/140989728
