Title :
The Moser-Tardos Framework with Partial Resampling
Author :
Harris, David G. ; Srinivasan, A.
Author_Institution :
Univ. of Maryland, College Park, MD, USA
Abstract :
The resampling algorithm of Moser & Tardos is a powerful approach to develop versions of the Lovasz Local Lemma. We develop a partial resampling approach motivated by this methodology: when a bad event holds, we resample an appropriately-random subset of the set of variables that define this event, rather than the entire set as in Moser & Tardos. This leads to several improved algorithmic applications in scheduling, graph transversals, packet routing etc. For instance, we improve the approximation ratio of a generalized D-dimensional scheduling problem studied by Azar & Epstein from O(D) to O(log D/ log log D), and settle a conjecture of Szabo & Tardos on graph transversals asymptotically.
Keywords :
graph theory; scheduling; Lovasz local lemma; Moser-Tardos framework; Szabo & Tardos conjecture; approximation ratio; generalized D-dimensional scheduling problem; graph transversal; partial resampling algorithm; Algorithm design and analysis; Educational institutions; Random variables; Routing; Silicon; Standards; Vectors; Lovasz Local Lemma; approximation algorithms; graph transversals; independent transversals; packet routing; randomized algorithms;
Conference_Titel :
Foundations of Computer Science (FOCS), 2013 IEEE 54th Annual Symposium on
Conference_Location :
Berkeley, CA
DOI :
10.1109/FOCS.2013.57